2026-10-14 - Flock Presentation

Summary

This is a public informational presentation on Flock Safety automated license plate readers (ALPRs), given by Anthony Gribbons of the Charlton Police Department with the police chief present and Carl Bock of Flock Safety joining by Zoom. It is not a Select Board vote; it was held ahead of a non-binding vote on the question that residents refer to as coming "in a week or so". Summary built from ASR auto-captions (no speaker labels, names often misspelled; the town name is frequently rendered "Charleston", "Chelmsford" or "Charlotte").

Lt. Gribbons explained what the system captures: still photos (not video) of vehicles on public ways, with plate, make, color, body type and distinguishing features such as roof racks or bumper stickers, plus GPS location, timestamp and direction of travel. He said there is no facial or voice recognition, no driver or owner information (officers must separately run the plate through CJIS/RMV), and no traffic ticketing. Data is owned by the town, auto-deleted after 30 days, and every search is permanently logged with a case number and reason. The department has four fixed solar-powered cameras (Route 169 on Southbridge Road, Main Street at the library, Stafford Street at Route 20, and Stafford Street at Northside Road), installed between fall 2024 and early 2025 under a 2023 agreement, funded from the police budget at about $3,000 per camera per year. An agreement to buy four more cameras is on hold. He said that immigration and reproductive-health searches are blocked, that sharing is approved agency by agency (about 1,109 agreements, plate-only searches), and that he audits searches monthly with an AI-assist flag for irregular patterns and gives the results to the deputy chief. He described successes: a Worcester double-homicide suspect, missing elderly people, a home-equity-line check fraud ring traced from Maine to Georgia, bank robberies, tire dumping, stolen cars (including a stolen-car "cool down" site), hit-and-runs and a false insurance claim. He also covered the limits: nothing is retained beyond 30 days unless downloaded as evidence, so unreported crimes can go unsolved.

Residents pressed hard on privacy and trust: third-party facial-recognition integrations, data being sold or breached (a Connecticut/Ledyard breach article and reports of hacked cameras), federal or ICE access and subpoenas, the 1,100+ sharing partners and private-network access, accuracy and misreads (one resident cited a family held at gunpoint over a misread plate), a federal Oklahoma ruling against ALPRs, Fourth Amendment concerns, cameras reported in parks and playgrounds, the "OS Investigate" product, and how to keep future police leadership from adopting more invasive features. Carl Bock said Flock does not and will not do facial recognition on these cameras, that the AWS GovCloud was never breached (though individual cameras have been accessed), that a third-party penetration test and a bug bounty exist, and that the Falcon cameras cannot search for people. Police said the Select Board could override or act on a non-binding vote, that purchases fall within the chief's budget authority, and that other towns were reported to have suspended or dropped Flock (residents cited 30-plus towns). Lt. Gribbons ended with a live demonstration of the Flock search interface (a red-sedan search narrowed from about 1,400 results to six by adding stickers), year-to-date figures (about 5 million reads, 23,000 reads that day, 386 stolen vehicles and 129 sex-offender hits via NCIC), and an audit/AI-flag demo.

Issues discussed: Flock Automated License Plate Readers

Transcript

00:04 | All right. So, what we're here for is
00:06 | first of all, my name is Anthony
00:07 | Robbins. I'm a lieutenant of the police
00:08 | department.
00:09 | Um
00:10 | and I'm here tonight to try and give you
00:12 | a little bit of
00:14 | information about Flock, how we use
00:15 | Flock,
00:17 | and
00:18 | then we'll answer some questions.
00:20 | Um I would ask just uh Mr. Bock is on
00:23 | here from Flock.
00:25 | Um if you have any questions for him
00:27 | directed to Flock,
00:29 | he can ask those right at the end of the
00:31 | presentation so I can He's got other
00:33 | commitments and I can get him
00:35 | moving along and then I'll stay as long
00:37 | as I need to to
00:38 | answer whatever you have and demonstrate
00:40 | whatever you need to demonstrate. Does
00:41 | that work for you, Carl?
00:45 | >> That sounds great. Thanks.
00:46 | >> Okay.
00:49 | So, what is an automated uh license
00:52 | plate reader? So,
00:54 | they're cameras that are specifically
00:56 | designed.
00:57 | Um
00:58 | they have an optical character
01:00 | recognition software in there that
01:02 | recognizes license plates.
01:04 | Um
01:06 | there are many forms of it and traveling
01:08 | all over our license plate readers are
01:09 | fixed fixed mounted. They're static to a
01:12 | pole. Um
01:15 | and our vendor is Flock Safety.
01:18 | So, Flock cameras photograph vehicles
01:20 | that pass the camera. They log the plate
01:22 | number, an image of the vehicle, GPS
01:24 | coordinates of where the camera was at
01:27 | the time, and apply a timestamp,
01:30 | and then put that into a searchable
01:32 | database.
01:34 | Uh software from Flock runs
01:36 | like I said, OCR on the plate, uploads
01:39 | the record to the Flock's cloud within
01:40 | seconds. And each record contains the
01:43 | plate number, a full frame vehicle
01:44 | photo, the camera's GPS coordinates, and
01:48 | the time of the uh
01:50 | the date and time of the image.
01:53 | A lot of people see cameras around town,
01:56 | see cameras in general, think they're a
01:57 | Flock camera. This is what a Flock
01:59 | camera looks like.
02:00 | Um
02:01 | So, I know we've had some other
02:03 | inquiries about cameras.
02:05 | This one is
02:06 | what we use, and this is what it looks
02:08 | like.
02:09 | Um they're solar-powered, they have a
02:10 | battery backup.
02:12 | They sit about 12 to 14 ft off the
02:14 | ground.
02:15 | Uh
02:16 | Ours are all ins- I'm sorry. Three out
02:19 | of the four of ours are installed on
02:20 | independent poles.
02:21 | One is installed on a shared pole.
02:24 | Uh they're used for investigatory
02:26 | purposes, and they provide us with
02:27 | precise data regarding vehicles.
02:30 | Uh we find it to be a force multiplier.
02:32 | It allows us
02:34 | uh to do more and do more faster.
02:38 | >> Is is that what's on the uh police
02:40 | vehicles as well, or is that different?
02:42 | >> We don't have any cameras on the police
02:44 | vehicles.
02:45 | >> Oh, okay.
02:46 | >> So, we have nothing on our vehicles at
02:48 | all. I just We have four cameras,
02:50 | and they are mounted to poles.
02:52 | >> So, no other surveillance anything on
02:54 | those cameras.
02:55 | >> We don't have We have body cameras, but
02:57 | we have no cruiser cameras, and we have
02:59 | no other plate reader technology.
03:02 | So.
03:03 | What does a Flock uh license plate
03:06 | reader collect? Um it gets the license
03:08 | plate number, full and partial mar- uh
03:10 | matches, uh a full frame photograph of
03:13 | the vehicle, camera GPS coordinates,
03:16 | a time stamp accurate to the second,
03:18 | vehicle color, make, body type,
03:20 | and other attributes, roof racks, bumper
03:23 | stickers, image patterns, missing
03:24 | hubcaps,
03:25 | >> cats, spoilers,
03:26 | >> and a direction of travel.
03:28 | Um
03:30 | If a camera is pointed in a direction
03:31 | and the car is going in
03:33 | one direction, we know the direction of
03:35 | travel. We don't know where it's going,
03:37 | where it's come from, but we know that
03:38 | it's either
03:40 | uh going north or south if it was in
03:42 | front of the library.
03:47 | So, what Flock does and doesn't do.
03:49 | Um
03:50 | A Flock
03:52 | camera takes a point in time picture of
03:54 | a vehicle traveling on a public road.
03:56 | All of our cameras are on public ways in
03:58 | the public capturing vehicles in the
04:01 | public on the travel.
04:03 | Uh on the travel portion of the way, so
04:04 | that on the road.
04:06 | Uh they capture the make, model, license
04:08 | plate.
04:09 | They support only criminal
04:10 | investigations or other our
04:12 | investigations, missing people, crimes,
04:15 | what have you.
04:16 | Um
04:17 | and they keep permanently a record of
04:19 | every search made by a member of law
04:21 | enforcement. So, they dump their data
04:25 | every 30 days,
04:26 | but the search search is conducted by
04:29 | police officers uh
04:32 | are are retained permanently.
04:35 | Um
04:36 | what they don't do.
04:37 | They have no facial or voice recognition
04:40 | capability.
04:41 | Uh they don't allow for the searching of
04:43 | a person, a pet, clothing type.
04:46 | Uh they don't connect to the registry uh
04:49 | or contain any information about the
04:51 | driver. Uh
04:53 | and that's what say uh I apologize.
04:54 | Driver of the vehicle. Uh they also
04:56 | don't give us the
04:58 | you know, the year of the vehicle. They
05:00 | don't give us other, you know, sub
05:02 | models, what have you. Uh
05:04 | it generates it does not generate
05:07 | automatic traffic enforcement tickets
05:09 | and it does not track the vehicle
05:10 | continuously.
05:11 | Um
05:12 | >> Is the camera just constantly like when
05:15 | you're on the pike?
05:16 | Uh
05:17 | >> So, there
05:18 | there is an
05:19 | Carl, just correct me on this. There's
05:20 | an algorithm built into the camera where
05:23 | it's not taking video, it's taking
05:24 | stills in a
05:26 | >> Actual still pictures.
05:27 | >> still pictures, right. There are no
05:28 | videos.
05:29 | >> So, it must
05:31 | happen when it
05:33 | So, when a vehicle is nearing. Like it's
05:36 | not
05:37 | >> taking It It can be motion or it can be
05:39 | a uh like a six I believe it Carl,
05:42 | correct me if I'm wrong. It works out to
05:44 | six stills per second or so.
05:45 | >> Okay.
05:47 | >> Yeah, so it's it'd be motion activated
05:49 | when it senses the presence of a vehicle
05:50 | and taking still photos.
05:52 | >> Thank you.
05:56 | >> So, assurances that Flock is made to us
06:00 | um
06:02 | for privacy protection, the data is 100%
06:05 | owned by the customer, in that case the
06:06 | town of Charlton.
06:07 | Um and will never be sold. Private
06:10 | customers don't have access to
06:12 | government data or hot lists, and we
06:13 | don't have access to their
06:15 | data as well.
06:16 | Okay?
06:17 | Um
06:19 | the Flock automatically deletes data
06:20 | after 30 days by default. And a Flock uh
06:24 | LPR only takes images of publicly
06:26 | available information. So, the courts
06:28 | have said that a vehicle on a public way
06:30 | is in the public view, and that's
06:33 | where we have our cameras positioned,
06:35 | and that's where we take pictures.
06:37 | Um
06:38 | and only the the agency decides uh who
06:41 | to share with, not Flock. And again,
06:43 | there's no facial recognition, and they
06:45 | assure us of that.
06:46 | >> But, what happens when they integrate
06:48 | with other companies that are offering
06:49 | that or about to offer that?
06:51 | >> Flock is its own independent company, it
06:53 | does not
06:54 | >> the data, though. They make it access to
06:56 | other They make it accessible to other
06:57 | companies.
06:58 | >> Carl, can you hear that question?
07:01 | >> Novimoto, for example, has promised to
07:03 | do facial recognition as soon as you let
07:05 | them or as soon as somebody's
07:06 | interested, wants to pay for it. I don't
07:09 | know if they're going to make it active,
07:10 | but
07:10 | >> Can you hear his question, Carl?
07:12 | >> Yeah, related to facial recognition,
07:14 | it's not something that we make in any
07:16 | of our devices. It's honestly not
07:18 | something our customers are asking for.
07:20 | Um so, it's not anything that we have or
07:23 | plan to do.
07:23 | >> Well, you know, I know there's a company
07:24 | actively trying to sell that service to
07:27 | uh different police departments across
07:29 | the country saying that they're going to
07:30 | do it. So, what's the difference if one
07:32 | person does it, two people do it? I
07:33 | don't care if it's Flock. If somebody
07:35 | else can do it, it's the same thing.
07:36 | >> We don't subscribe to them, and we
07:38 | wouldn't Would you agree, Chief, we
07:39 | wouldn't have that data?
07:41 | >> No, but it all goes to Flock. It's not
07:42 | You don't get access to it. No, that's
07:44 | That's a
07:45 | >> It's a hindrance to you cuz you can't
07:47 | see all of it, but they have the access.
07:49 | They can sell the data to whoever they
07:50 | want.
07:51 | >> They don't They assure us they don't
07:52 | sell that data and they don't
07:54 | They give us assurances in the contract
07:56 | that they don't do that.
07:58 | And that would be a breach of contract
07:59 | and we'd have a an action against them.
08:01 | >> Well, how are you ever going to know?
08:04 | You're not going to see it happening in
08:05 | San Diego. You're not going to have any
08:06 | You're not You're You You have other
08:07 | things to do.
08:08 | You're not going to
08:09 | >> Well, I mean, it Word travels fast on
08:11 | this stuff if
08:13 | if it if it came out
08:14 | >> enough cuz you don't seem to know about
08:15 | what I'm talking about.
08:16 | I just picked up the data off the
08:17 | internet 10 seconds ago and we put
08:19 | >> You got to be careful what you read
08:20 | about what you see on the internet, too.
08:21 | >> Oh, but these are legitimate news
08:22 | sources that have that
08:24 | >> That's
08:24 | >> Oh, absolutely. There's a bunch of
08:25 | garbage on the internet.
08:26 | >> I I would tell you this, though, if the
08:28 | data of Flock was breached, you'd get a
08:31 | picture of your license plate. Same
08:32 | thing I can walk out and take a picture
08:34 | of your car right now. Might show me
08:36 | going down the street, but it's not a
08:37 | person. It doesn't track a person. It
08:39 | tracks a vehicle type.
08:41 | >> Yeah, but will you promise to never tell
08:43 | anyone
08:43 | >> China would do with my license plate.
08:45 | >> Would you promise to never use that
08:47 | technology
08:47 | >> I don't know what China would do with my
08:48 | license plate. I don't know what China's
08:50 | going to do with that information.
08:51 | >> Well, they could do all sorts of stuff
08:53 | with your information.
08:54 | >> Well, I mean, what's to prevent
08:55 | >> happens when facial recognition
08:57 | recognition comes in? Japan starts to
08:58 | use it.
08:59 | >> I promise not to use it. Do you?
09:01 | >> Well, let's get Let's get that codified.
09:03 | >> It's not something we're interested in.
09:04 | >> It It Flock
09:05 | already uses facial recognition. It's
09:07 | fed into AI systems everywhere and you
09:10 | can absolutely identify people by what
09:12 | in the vehicle just through the ALPRs.
09:15 | >> Let me work through the PowerPoint and
09:16 | then we'll address questions. Okay?
09:19 | Um
09:20 | So,
09:21 | Flock data is encrypted in transit and
09:23 | it rests
09:25 | um in a line with the FBI uh
09:27 | uh CJIS uh system.
09:30 | It's uh
09:31 | the
09:32 | aligns with the National Defense
09:33 | Authorization Act. It's independently
09:36 | audited for breaches, Um and everything
09:39 | is aligned with international standards
09:42 | for data security, and
09:45 | there's a higher education community
09:47 | vendor assessment tool.
09:49 | Flock is also aligned with security
09:51 | protocols established by
09:55 | the NIST cybersecurity as well as the
09:56 | cloud security
09:58 | CAIQ framework, and their data is stored
10:01 | on the AWS government cloud.
10:07 | Motivators for police departments to use
10:09 | ALPRs: crime prevention, developing
10:11 | investigative intelligence, locating
10:14 | missing or endangered persons,
10:16 | recovering stolen vehicles,
10:18 | enhancing public safety through
10:19 | vehicle-based intelligence,
10:21 | increasing officer efficiency and to act
10:23 | as a force multiplier,
10:25 | to exonerate innocent persons, provide
10:28 | real-time alerts, and identify vehicles
10:30 | associated with wanted persons or
10:32 | criminal activities when they're in
10:33 | trouble.
10:35 | The addition of cameras in Charlton. The
10:38 | Charlton Police Department followed
10:39 | national trends regarding the addition
10:41 | of body-worn cameras and ALPRs
10:43 | in locating suspects or missing and
10:45 | endangered persons. It's estimated that
10:48 | approximately 70% of all investigations,
10:51 | whether a missing person, whether a
10:53 | criminal act involved a motor vehicle.
10:55 | Video evidence assists in fair
10:57 | prosecutions, police accountability,
11:00 | ALPR devices allow for the police to
11:03 | focus on real-time data.
11:05 | Photographic evidence and Flock images
11:07 | presented to jurors allow jurors the
11:08 | opportunity to examine the information
11:11 | that led police to making a conclusion
11:13 | regarding an investigation.
11:15 | Flock images also assist in the
11:17 | exclusion of persons unrelated to the
11:19 | investigation of an event.
11:24 | Why do we implement ALPRs?
11:28 | Starting increase in criminal activity
11:30 | and mental health events between 2020
11:32 | and 2023. Couple examples, bank
11:35 | robberies.
11:36 | Uh we had two bank robberies. We had
11:38 | several commercial, residential, and uh
11:41 | and commercial breaking and entering.
11:43 | We've had multiple motor vehicle thefts.
11:45 | We've had breaking and enterings into
11:46 | motor motor vehicles.
11:48 | We've had hit-and-run crashes with
11:49 | serious personal injury and property
11:51 | damage.
11:52 | We've seen an increase in mental
11:53 | health-related calls where people were
11:54 | reported to be in motor vehicles.
11:57 | We've also seen a regional crime trend
11:59 | uh
12:00 | where residents of Sharon were
12:01 | victimized or the individuals
12:03 | responsible for committing the crimes
12:05 | uh had ties to the area or traveling
12:07 | through Sharon in motor vehicles.
12:09 | Um these investigations um resulted in
12:12 | the identification and prosecution of
12:14 | several suspects
12:15 | and locating several persons in crisis.
12:17 | However, officers spent hundreds of
12:19 | hours investigating these events,
12:21 | eliminating uninvolved parties, and
12:23 | identifying evidence prior to the actual
12:25 | apprehension of suspects or locating an
12:27 | individual in crisis.
12:29 | In several of these instances, the loss
12:31 | of time impacted the investigation or
12:33 | locating the person
12:34 | who needed assistance.
12:39 | As a result of these investigations, the
12:41 | question was asked, "Can we give our
12:43 | officers a better tool to help prevent
12:44 | and solve serious crimes and locate
12:46 | individuals in crisis?"
12:48 | Um
12:49 | stationary cameras, everybody has them.
12:51 | Ring cameras.
12:53 | Uh um
12:54 | you know, you have a people have cameras
12:56 | fixed to their houses, businesses have
12:57 | cameras. There's a lot cameras
12:59 | everywhere we go in society. Um but a
13:02 | stationary camera without ALPR
13:04 | technology
13:06 | um
13:06 | we found them
13:08 | they were beneficial that they gave us a
13:10 | picture of a car, but they didn't give
13:11 | us anything else and made us reactive
13:13 | rather than a proactive uh police
13:15 | department.
13:16 | Uh generally, the crime occurred first.
13:19 | We then had to determine what camera
13:21 | might have caught uh captured something.
13:23 | And then locate the footage, review it,
13:25 | and try to identify a vehicle or other
13:27 | piece of information to advance our
13:29 | investigation.
13:31 | We were looking for technology
13:34 | that could efficiently help us identify
13:36 | a vehicle associated with a crime and
13:37 | determine where that vehicle may have
13:40 | been at a specific point in time. We
13:42 | also wanted the ability, when we had a
13:44 | specific vehicle connected with a
13:45 | legitimate law enforcement concern, such
13:47 | as a stolen car, a wanted person, a
13:49 | missing person, or an active
13:51 | investigation, to receive information
13:53 | that could help our officers know it was
13:54 | in town and locate it or determine where
13:57 | it was possibly headed.
14:00 | This would assist us in directing
14:01 | resources to prevent another criminal
14:03 | act or to locate a person in crisis so
14:06 | they can receive medical care or mental
14:08 | health services in a more timely manner.
14:13 | The capabilities of Flock provided this
14:15 | tool with the safeguards we felt were
14:17 | necessary regarding the information
14:18 | captured, the network capabilities, the
14:21 | data retention, and the customer
14:22 | service.
14:26 | We understand that with greater
14:27 | capacity, there need to be greater
14:29 | safeguards. So, system audits. The
14:32 | system can be audited at any time
14:33 | showing exactly what's searched and by
14:35 | whom, both within the department and
14:38 | searches conducted by other agencies.
14:40 | Retention rules. Flock data is
14:43 | secured for a maximum of 30 days,
14:45 | with the data being owned by the
14:46 | Charlton Police Department and stored on
14:48 | the Flock network.
14:50 | We have it set up so that there's no
14:51 | searches for immigration or reproductive
14:53 | rights issues. In some states, these are
14:55 | crimes.
14:58 | We don't allow searches for those
14:59 | reasons. They are excluded searches.
15:02 | Um we did this based on some positions
15:05 | from the ACLU in accordance with the
15:07 | Massachusetts Shield Law. Um
15:10 | and then limit data sharing. We limit
15:12 | our data sharing to agencies with a an
15:14 | investigative purpose in town.
15:17 | We don't just
15:19 | it's like group select agencies. We
15:21 | select
15:22 | for a purpose.
15:25 | >> How do you stop ICE from getting that?
15:27 | And
15:28 | what if the feds just want to issue
15:31 | >> So, if the
15:32 | >> request for that. I know you said you're
15:33 | not sharing it, but
15:34 | >> We don't share it.
15:34 | >> What stops anybody else from just
15:36 | logging onto cameras and looking?
15:37 | >> Because they are
15:39 | we know who is searching, the reason for
15:42 | their search, and it's identifiable back
15:45 | to a person.
15:46 | >> So, if the federal government could
15:47 | require it, and then
15:49 | there's no way to stop it, and we have
15:50 | cameras everywhere.
15:52 | Regardless. Like you can't You're not
15:53 | going to
15:54 | >> Well, I mean, if the federal
15:55 | >> If they ask for it, but they ask for all
15:56 | of it. Like you have
15:57 | >> Let me Let me finish this, and then
15:59 | we'll go back to that. But I'll
16:00 | >> All right.
16:00 | >> I'll keep it.
16:02 | So, um terms of service and data
16:05 | ownership. As part of the protections to
16:07 | ensure that the data belonging to the
16:08 | Charlton Police Department is properly
16:10 | used,
16:11 | um
16:12 | we've ensured that Flock uses the
16:14 | information, and it's included in the
16:16 | that it shall
16:17 | uh
16:18 | not include any sharing by Flock of any
16:20 | customer data.
16:22 | Flock does not own
16:23 | um and shall not sell any of our
16:25 | customer data.
16:26 | At the end of the retention period,
16:28 | Flock will delete all customer data.
16:31 | And civil civil, immigration, and
16:33 | reproductive health crimes in other
16:34 | states shall not constitute a risk of
16:36 | imminent harm, and are excluded as
16:38 | search reasons through the Charlton
16:40 | uh interface.
16:44 | Privacy protection should not depend on
16:46 | assurances and a good policy and
16:48 | procedure. We understand that. We have a
16:50 | policy um
16:52 | that was created
16:54 | directly at Flock and ALPRs for the town
16:56 | of Charlton.
16:57 | Previously, we treated Flock data as
17:00 | CJIS data, um which is a higher level of
17:03 | protection, um
17:05 | but it it missed some of the stuff that
17:07 | ALPRs
17:09 | uh are specific to. So, we have a CJIS
17:12 | policy that's still in effect, and we
17:14 | have an ALPR policy.
17:16 | Individual accountability of each
17:17 | officer. Officers are subject to
17:19 | discipline for abuse based on
17:21 | circumstances of the violation. Every
17:23 | officer, anybody that searches Flock.
17:26 | There's a code. Um
17:29 | it's assigned to every single search. We
17:31 | know who searched it, which agency they
17:33 | searched from, the time and date of the
17:35 | search, and what they searched.
17:38 | We conduct internal audits of the
17:39 | system.
17:40 | Uh we use it the AI assist function to
17:44 | determine other patterns. Um
17:46 | we've just gotten into that uh that
17:48 | feature.
17:49 | But we manually audit the the searches
17:53 | that are conducted.
17:54 | Um
17:55 | and then we have technology controls and
17:57 | contractual restrictions. Um
18:01 | we limit searches for certain for plates
18:04 | only. Um
18:06 | we limit searches um
18:10 | so that somebody has to put in a call
18:12 | number or and a reason why they're
18:15 | searching.
18:19 | So we've had a lot of Flock successes in
18:21 | Charlton.
18:22 | Um Michelle, one of the times she came
18:25 | in to speak to me, had asked and I have
18:28 | not gone through and just compiled a
18:29 | list. This took me a few days. It the
18:32 | chief can tell you I
18:34 | live with this now. But um
18:37 | you know, we put the the Flock cameras
18:39 | out and
18:41 | it was mentioned in a a training um that
18:44 | we have Flock.
18:45 | Within
18:47 | it was days of us uh having Flock
18:49 | cameras.
18:50 | Um
18:52 | we used the cameras to identify a double
18:54 | homicide suspects in Worcester who were
18:56 | believed to be traveling on Stafford
18:58 | Street back to New Jer- uh Jersey and
18:59 | New York.
19:00 | Um
19:02 | we quickly found a resident uh with
19:04 | dementia who had left their residence in
19:06 | a vehicle uh without a phone.
19:09 | Uh we have officers that have worked
19:11 | with the deep that work with the DEA.
19:13 | Um,
19:14 | and there was a very high-profile
19:16 | recent, um,
19:19 | drug seizure and drug prosecution at the
19:21 | federal level that we assisted with that
19:23 | had ties to Charlton in the area. Um,
19:27 | we assisted the Mass State Police,
19:29 | um,
19:30 | in ruling out that somebody came through
19:32 | Charlton, so
19:33 | uh, following homicide. So,
19:36 | like I said before, we can identify a
19:38 | suspect and we can also rule somebody
19:40 | out. Um,
19:41 | and that's why I included some of these
19:42 | things.
19:43 | Um,
19:44 | we've located people related to, uh,
19:48 | thefts in storage areas,
19:50 | um,
19:52 | identified suspicious, uh, suspicious
19:54 | people who were believed to be
19:55 | surveilling a residence for a breaking
19:57 | and entering.
19:59 | Um,
20:00 | we assisted in the identification of
20:02 | vehicles
20:03 | and suspects involved in fraud where the
20:05 | victims were defrauded of $21,000.
20:08 | Um, but you can I can just keep going
20:11 | through this if you want to see, but
20:12 | like,
20:13 | hit and runs,
20:15 | um,
20:16 | if we look at dumpings that have
20:17 | happened at the town cemetery on Route
20:19 | 20,
20:20 | um,
20:21 | the recovery of stolen vehicles.
20:25 | Several property crimes have been
20:27 | solved. Um, again, assisting other, uh,
20:31 | agencies. Again, illegal tire dumping on
20:33 | several occasions.
20:35 | Um, you might not think that tire
20:36 | dumping is a big deal. I don't know if
20:38 | anybody saw the stuff at Zorba's and
20:40 | some of the other places.
20:41 | Uh, tires cost
20:43 | tens of thousands of dollars to remove.
20:46 | Um, and we were able to identify
20:47 | somebody off of that.
20:49 | Um,
20:51 | again, hit and runs, breaking and
20:52 | enterings, stolen vehicles.
20:55 | And these are just
20:57 | a snapshot of what we've used it for.
20:59 | Um,
21:01 | this is a a big one, um, in my opinion,
21:03 | in winter 2026. So, when you
21:07 | own a home or you have a mortgage or a
21:09 | home equity line,
21:11 | a lot of your data is
21:13 | open it's a public record.
21:15 | Um there are people
21:18 | and criminal organizations that go in
21:20 | and mine that data
21:23 | and they were writing checks against
21:24 | people they were
21:26 | timing out the payments that people had
21:27 | made, so they knew that on a home equity
21:30 | line that people had
21:32 | I have a $50,000 home equity line, the
21:33 | person's had it for so many months,
21:35 | they've paid it down
21:37 | you know, to $40,000
21:39 | and they were writing checks on the
21:41 | remaining amount that was left the line
21:43 | of credit. Um we had several of those
21:46 | here in town, several in Oxford
21:49 | and when we put everything together um
21:52 | we were able to trace from Maine to
21:54 | Georgia
21:55 | um where this crew had had defrauded
21:58 | people of of
22:00 | a significant amount of money. It to the
22:01 | point where we still haven't figured out
22:04 | exactly the total amount. It's that
22:06 | high.
22:07 | Um
22:09 | >> How did the cameras
22:11 | How did the cameras help? I mean
22:12 | >> So, the person drove by the library
22:13 | before they went to Cornerstone Bank. Um
22:16 | and they were wearing they were changing
22:18 | up their clothes, they were changing up
22:19 | their disguise, but a person remembered
22:21 | the car. Um the bank camera didn't
22:23 | capture
22:25 | the best image, the flock camera
22:26 | captured the best image of the car.
22:28 | Um and they were going from an Oxford
22:31 | branch to a Charlton branch and that's
22:33 | how they were uh
22:35 | identified.
22:37 | Um we again hit and runs, another
22:40 | missing elderly person.
22:44 | Um
22:46 | in the spring of this year
22:48 | um there was a site in Charlton that was
22:50 | used as what they call a cool down
22:51 | location for stolen cars.
22:53 | Um because of our location
22:56 | and the roads that surround us and go
22:58 | through us um we tend to be a hotspot
23:01 | for stolen cars. What they were doing is
23:04 | putting a car in a parking lot, letting
23:07 | it sit for a bit to see if it had uh
23:10 | telematics that if they deactivate the
23:12 | telematics in the car. And they were
23:15 | putting them in parking lots to wait and
23:16 | see if their telematics
23:18 | deactivations had worked before they
23:20 | moved the car somewhere else.
23:22 | Um
23:24 | we've had North Carolina
23:26 | uh
23:26 | in the spring of this year report to us
23:28 | a lot of stolen vehicle activity
23:30 | uh in Charleston.
23:32 | And
23:33 | um we've seen it used in investigations
23:36 | for an assault and battery with a
23:37 | dangerous weapon, several domestic
23:39 | violence cases. Um and um a stolen
23:43 | vehicle investigation from a commercial
23:45 | property with multi-state ties.
23:53 | Again, I did them up until fall of this
23:55 | year. Most recently, um we used it
23:58 | um a party that had made a terroristic
24:00 | threat to a facility here in Charleston.
24:03 | Um when we put the plates in the in the
24:06 | Flock system, we're able to show that
24:08 | the person was in town the night before.
24:11 | Um
24:12 | but we've got some highlights here.
24:15 | A lot of it is
24:18 | basic stuff with
24:21 | quality of life issues. Somebody
24:22 | dumping, somebody, you know, illegally
24:24 | dumping, somebody stealing somebody's
24:26 | property, somebody stealing somebody's
24:28 | money.
24:29 | Um
24:30 | and identifying and locating people
24:33 | needing help.
24:38 | So, these aren't just hypothetical uh
24:40 | hypothetical examples of what the
24:42 | technology might accomplish. These are
24:44 | real successes that the uh Charleston
24:46 | Police Department has had where we've
24:48 | solved crimes and recovered and arrested
24:50 | people.
24:51 | Um they resulted in Charleston residents
24:53 | and business owners being able to seek
24:55 | restitution for property crimes. and
24:57 | they've assisted in getting uh persons
24:59 | in crisis
25:00 | uh to services to assist them.
25:04 | This is the nuts and bolts of it, okay?
25:05 | This is my cruiser.
25:07 | Um what happens when an officer checks a
25:09 | plate? This is the image that you get,
25:12 | okay?
25:13 | Um I personally like this image
25:15 | only because certain
25:18 | Has anybody ever tried to use an OCR
25:19 | scanner on a piece of paper?
25:21 | Sometimes it'll juxtapose numbers,
25:23 | letters.
25:24 | My license plate has an eight and a B
25:26 | next to each other. It's able to
25:27 | identify and separate out the eight and
25:29 | the B.
25:30 | Fives on a
25:32 | on the Xerox copier will sometimes come
25:34 | up as S's, okay? It identifies a five
25:37 | and two more eights, and it does it
25:39 | successfully.
25:41 | I didn't pick up on that until about 2
25:43 | weeks ago when I was
25:44 | >> Do you know the rate that it picks up
25:46 | successful?
25:47 | >> We've never had a miss. We've never had
25:49 | a a misread.
25:52 | >> Out of
25:53 | millions of photos?
25:54 | >> I would say more than a half a million,
25:56 | if not more.
25:58 | >> There was a recent misread where a
25:59 | family was yanked from their car and
26:01 | held at gunpoint on the ground because
26:05 | an eight was a B.
26:06 | >> In Charlotte?
26:07 | >> No, not in Charlotte. But millions of
26:08 | reads and no misreads just in Charlotte.
26:11 | >> We've had no misreads in Charlotte.
26:13 | >> How do you know you don't have misreads?
26:14 | >> They pre-checked them all.
26:15 | >> Every time we've checked one, we've
26:17 | never had a misread.
26:18 | >> But that's only in the times you
26:19 | checked. I mean
26:20 | >> We don't have a an ability to do it
26:22 | otherwise.
26:23 | So,
26:26 | So, this is
26:27 | if we were if my cruiser is involved in
26:29 | a crime, this is what we would get,
26:32 | okay? So, it's a picture of the car, a
26:34 | picture of the license plate, and the
26:35 | location with a time and date stamp uh
26:38 | of where the camera was that captured
26:40 | it. If you look,
26:42 | it gives you the license plate. It tells
26:44 | you that it's an SUV, that it's a Chevy,
26:46 | and it's either silver or gray, time
26:48 | stamp, organization that
26:50 | uh looked for it. It'll look up top and
26:53 | it see my name at the top. So, it
26:55 | identifies me as the person searching
26:57 | it.
26:58 | I did this for training demonstration,
27:00 | so I gave the
27:01 | the search reasons and the case number.
27:03 | Um,
27:04 | but again, it's a device that LPR, grid
27:06 | coordinates, and a physical address.
27:11 | So, when an image is captured, the
27:13 | officer then must run the plate number
27:14 | through the CJIS system and the RMV to
27:16 | find out who the car is registered to.
27:18 | There is nothing in Flock that gives us
27:20 | who the car is registered to or any
27:22 | other details.
27:23 | Um,
27:24 | >> Except a
27:25 | just a search button, right?
27:27 | What do you mean? Just so you have to to
27:28 | find out the who owns it, you have to
27:30 | search it up. So, you have to type it in
27:32 | manually or
27:33 | >> We type it in manually.
27:34 | >> That's that's like six whole letters,
27:36 | right? It's not actually just hitting a
27:37 | button.
27:37 | >> system, though.
27:38 | >> system.
27:38 | >> Different system.
27:39 | >> That's also a system we've had for 40
27:41 | years.
27:42 | >> I mean, I've got CJIS. I'm not trying to
27:43 | accuse you guys of do of being bad or
27:46 | having bad or having bad intentions or
27:48 | whatever, right? Small town of Charlton,
27:50 | we're all cool, right? But
27:52 | it doesn't stop there. It's not that's
27:55 | not just you guys. You have great
27:56 | intentions in using this. We're nobody's
27:58 | saying you don't. You've done some great
28:00 | things with it, but
28:02 | where's it going? And you said you know
28:03 | I've committed no crimes. I mean, you
28:04 | violated my fourth amendment when you
28:06 | took a picture.
28:07 | >> I did violate your fourth amendment, but
28:08 | we'll discuss that later, okay?
28:09 | >> So, I I'm trying to hold them
28:11 | >> opposite. So, the CJIS and the RMV is
28:14 | also monitored and audited. So, every
28:17 | every search is is recorded in there as
28:19 | well. All right, officers must have a
28:21 | legitimate reason for using these
28:22 | systems. An officer will know um, who's
28:26 | won't know whose car
28:28 | it is unless they use the RMV. Um, these
28:31 | are the two ways to get a name
28:33 | associated with a plate. And then
28:35 | the hot list for people that have been
28:37 | entered as wanted for serious crimes,
28:41 | um,
28:42 | in NCIC,
28:44 | um, or for missing persons, which are
28:45 | also entered in NCIC.
28:48 | A Flock alert alone is not probable
28:50 | cause and requires an officer to verify
28:52 | the data and make independent
28:53 | observations what they see.
28:56 | If if we get a Flock alert that my car
28:59 | is wanted, we get behind a car, the
29:02 | officer
29:03 | is going to confirm that what they see
29:06 | on that plate matches what's on the
29:08 | Flock data. They're not going to make a
29:10 | stop just off the Flock alone.
29:15 | So, I saw this over the weekend and I
29:17 | added it. I don't know if anybody saw
29:18 | this on
29:20 | um
29:22 | on Facebook.
29:24 | There was a comment added to the to the
29:26 | Facebook post where the person had more
29:28 | information and I don't know how um
29:32 | than I had in CCHIS. Okay? And it was a
29:35 | publicly searchable they said it was
29:38 | publicly searchable. So, I Googled the
29:40 | license plate
29:41 | and I was able to find out a a lot of
29:43 | stuff about the motor, about um the
29:46 | mileage on the car, um it kind of the
29:49 | vehicle identification number,
29:51 | um
29:52 | and
29:53 | it kind of was
29:56 | an eye-opener to me as to what is
29:58 | publicly out there regarding your stuff.
30:01 | Um and it goes above what I have access
30:04 | to as a police officer. Um
30:06 | just off of a Google search.
30:09 | So, Flock audits. So, this is what an
30:11 | organizational audit looks like. Um
30:14 | if you
30:16 | uh
30:17 | search, um it'll tell you the reasons
30:20 | that we search, it'll tell you um who
30:24 | searched, it'll give you a breakdown of
30:26 | of times and dates of the searches as
30:28 | well.
30:29 | Oop.
30:30 | >> How
30:32 | real quick on the audits, how often
30:35 | are audits performed and are they
30:37 | standard or are they also a a audit, you
30:41 | know?
30:41 | >> So, we do an audit once a month because
30:42 | everything's retained.
30:43 | >> Okay.
30:44 | >> And then we do
30:46 | uh the AI assist will alert us to an
30:48 | irregular search pattern. And I can go
30:51 | into that one second.
30:52 | >> Okay.
30:53 | >> This is what an audit looks like. This
30:55 | is the ID for the search. Um
30:58 | the person who searched it, the network
30:59 | searched, date times, reason for the
31:02 | search.
31:03 | Um case numbers.
31:05 | Um I did block some of this stuff out
31:07 | because I knew we were being recorded
31:08 | for
31:09 | uh
31:10 | public access.
31:12 | Um it'll tell you the filters. It'll
31:14 | tell you uh the search time and the
31:17 | search type, okay? Um
31:19 | with regard to your question about
31:21 | audits.
31:23 | So, we have one audit that we identified
31:26 | and then
31:28 | it was an abnormality
31:29 | and
31:31 | the AI audit function also identified
31:33 | it. What it turned out to be, we looked
31:35 | into it, what it turned out to be is it
31:37 | was somebody wanted for another crime.
31:40 | Um
31:41 | they were believed to be traveling
31:43 | to Charlton from a neighboring
31:44 | community.
31:45 | Um an officer sitting on Route 20 where
31:47 | there's no flock cameras,
31:49 | um went to his
31:52 | uh MDT,
31:54 | the computer inside the cruiser, puts
31:56 | the
31:58 | uh license plate in there
31:59 | and continually hits refresh on Stafford
32:01 | Street because he believed, "Hey, the
32:03 | car
32:05 | laws of probability, the car's going to
32:07 | go down Route 20 or Stafford Street. I
32:08 | can sit on Route 20 and with my eyes
32:10 | look, but I can't be in two places at
32:12 | the same time and I can see if that car
32:14 | is going to go by
32:15 | um
32:17 | on Stafford Street."
32:18 | Because he kept hitting refresh every
32:20 | minute,
32:21 | um it identified a pattern that, "Hey,
32:23 | he's repetitively searching this plate.
32:25 | Why is he repetitively searching this
32:27 | plate?" And then we were able to see,
32:29 | "Oh, there was a an incident in another
32:31 | community
32:32 | uh related to that vehicle."
32:38 | So, we also
32:39 | retain these um
32:42 | It's the same information in the last
32:43 | screen, but in Excel spreadsheet so that
32:45 | we can retain this information, get it
32:48 | in a a usable format that
32:50 | um that we can use. The biggest use for
32:53 | this, in in all honesty, is that
32:56 | somebody misuses Flock.
32:58 | Okay? If an officer if we determine that
33:01 | an officer misused Flock, this is the
33:03 | evidence that we would
33:04 | uh discipline them or take action with.
33:07 | Um
33:09 | that's what this is is in there for.
33:14 | So, where we place cameras and why.
33:16 | Flock cameras are run on solar power
33:18 | utilizing a secure wireless connection.
33:21 | So, we need sunlight and we need a cell
33:23 | connection. There are certain areas of
33:25 | town that
33:26 | that's not possible. So, we can't we
33:28 | can't use them.
33:30 | All right? We base the location on where
33:32 | we had seen trends in travel following
33:34 | investigations, of course in accordance
33:37 | with building codes, other regulations,
33:39 | and where we could get cells uh
33:41 | consistent cell service, and sun.
33:44 | We have four cameras right now. Route
33:45 | 169 on Southbridge Road, Main Street in
33:48 | front of the library, Stafford Street at
33:50 | Route 20, and Stafford Street at
33:52 | Northside Road.
33:53 | Stafford Street is a very busy
33:56 | road for criminal activity. Just
33:59 | >> Are there a lot more that
34:01 | are going to be planned at this point?
34:03 | >> We have put a hold on everything right
34:05 | now, um but we did purchase we did enter
34:08 | into an agreement to purchase four more.
34:11 | Um but we have not installed them, we've
34:12 | not received them, and
34:14 | we agreed to put those on hold, and
34:16 | we hold true to our word.
34:17 | >> Curious if the the funding for the
34:19 | cameras, their upkeep, and the program
34:21 | comes from the police budget or the town
34:22 | budget?
34:23 | >> Comes from the police budget.
34:24 | >> Comes from the police budget.
34:25 | >> Yeah.
34:26 | >> When were the When How long have the
34:28 | cameras been installed?
34:30 | >> So, we entered into an agreement in '23.
34:33 | >> Okay.
34:33 | >> They were installed in
34:36 | uh
34:37 | between the fall of '24
34:40 | and the uh
34:42 | early part of '25. So,
34:45 | this is a video from Flock just to give
34:48 | you an idea of what Flock
34:49 | >> Ashley.
34:50 | I build software at Flock.
34:52 | You've probably heard of license plate
34:54 | readers and thought surveillance.
34:57 | I understand why it feels that way.
34:59 | But, that's not what we do.
35:01 | License plate readers take images of
35:03 | vehicles. Flock LPR is used after a
35:05 | crime happens.
35:07 | Things like stolen cars, hit-and-runs, a
35:10 | missing person.
35:12 | An officer can search for a specific
35:13 | vehicle tied to that case.
35:16 | The LPR system doesn't track people, and
35:18 | it doesn't predict when crime will
35:20 | happen. You can only search for vehicles
35:22 | on our LPR cameras.
35:24 | License plates, make, model, color, and
35:28 | other unique identifiers like bike
35:30 | racks, trailers, and toolboxes.
35:32 | There's no facial recognition and no
35:34 | ability to track a person. In most
35:37 | communities, the data automatically
35:38 | deletes after 30 days unless local laws
35:41 | require something different. The data
35:43 | collected doesn't live forever. Flock is
35:45 | built to help solve crimes faster and
35:48 | protect people's privacy. When you hear
35:50 | Flock LPR search, you may imagine some
35:53 | open, unrestricted database. Here's what
35:56 | the license plate reader search process
35:58 | actually looks like. The Flock system
36:00 | has built-in safety guardrails that
36:02 | limit who can access our tools.
36:04 | Law enforcement agencies also set rules
36:07 | based on their own policies. When an
36:09 | investigator is working on a specific
36:11 | case,
36:12 | say a kidnapping,
36:14 | they enter the offense type, case
36:16 | number,
36:17 | why they're searching, and narrow it
36:19 | down to a time and a place. Flock
36:21 | requires users to log why they're
36:23 | running an LPR search. If there's a
36:25 | match, the LPR search helps them figure
36:28 | out which car was there.
36:30 | And it does not give law enforcement any
36:32 | information on who was in that car,
36:35 | either the driver or any passengers.
36:37 | That's it.
36:38 | A focused tool to solve real cases.
36:41 | That means missing people come home
36:42 | faster,
36:44 | more justice for victims, and overall
36:46 | better safety for everyone. A fair
36:48 | question we get a lot.
36:50 | How do police chiefs, law enforcement
36:52 | leaders, and elected officials
36:55 | know what officers are doing in the
36:56 | Flock system? First, every search and
36:59 | sharing request is logged.
37:02 | Every search records who ran it, when,
37:05 | and why. Agency administrators can run
37:08 | audits to identify any searches that
37:10 | don't match normal casework.
37:12 | Agencies also own and control all of
37:14 | their data. Each law enforcement agency
37:16 | decides if and how they share their
37:19 | data, unless laws and policies dictate
37:21 | otherwise. Agencies can approve or deny
37:24 | data access requests from other law
37:26 | enforcement agencies on a one-on-one
37:28 | basis or by geographic radius. Some
37:30 | agencies will choose to join statewide
37:33 | or nationwide sharing networks where
37:35 | possible, and that's their choice.
37:37 | It's never auto-approved, and they can
37:39 | change the setting at any time. Flock is
37:41 | designed to provide evidence with
37:43 | transparency and accountability built in
37:46 | because we believe everyone has a right
37:48 | to be safe.
37:54 | >> Tell us if you've opted for statewide or
37:56 | nationwide or whom.
37:58 | >> We do selective, uh, we individually
38:01 | search, uh, individually approve every
38:03 | single one.
38:04 | >> Okay. So, there's no open sharing
38:06 | platform with any other agencies in the
38:07 | state?
38:09 | >> It's not open like
38:11 | >> We have 1,100 agreements open currently.
38:13 | >> 1,100 agreements.
38:14 | >> 1,109.
38:16 | >> You just basically said we don't.
38:19 | >> No, I said we don't we don't just open
38:21 | it up to everybody. We individually
38:23 | approve over the past 3 years requests.
38:27 | >> Sorry for
38:28 | I'll rephrase myself. Out of the 1,109,
38:32 | do they have open access or is each
38:34 | request individual for any data search?
38:38 | >> So they're limited we limit them to a
38:41 | uh
38:42 | vehicle search for plates only. No other
38:45 | analytic data or anything else.
38:47 | >> Such as what Charlton does itself, you
38:49 | said, yeah?
38:49 | >> I can show you that later. Um
38:52 | >> Just curious, sorry.
38:53 | >> No, that's fine.
38:53 | >> There's a lot to take in, so I questions
38:55 | are going to be difficult to follow.
38:57 | >> with you, that's why I handed out slips.
38:59 | If you if you have a At the end, I'll
39:01 | give you my email address if something
39:02 | comes up.
39:03 | You can
39:04 | >> Are those mostly local jurisdictions
39:07 | that we're talking
39:07 | >> A lot of them are local jurisdictions,
39:08 | but I'll explain it in a little bit.
39:10 | >> Okay.
39:11 | >> Um
39:12 | So the Charlton Police Department
39:14 | regularly assesses the value of the
39:15 | Flock product. Um and to see that it
39:18 | serves a legitimate public safety uh
39:21 | purpose here.
39:22 | We also ensure privacy protection,
39:24 | access limits, retention rules, and
39:26 | accountability standards are being met.
39:28 | Um
39:30 | We've seen that it's materially assisted
39:31 | us in in public safety objectives,
39:34 | solving crime, finding people.
39:37 | Um
39:38 | We've documented public safety outcomes
39:40 | um to justify the financial cost of the
39:42 | system.
39:44 | Um
39:45 | We've
39:46 | checked on policy, privacy issues,
39:49 | access, retention, and other violations.
39:52 | And if so, we're committed to taking
39:54 | corrective action on it.
39:56 | Um
39:57 | We've ensured that and checked to see
39:59 | that the limitation on data retention,
40:01 | information sharing, authorized access,
40:03 | and permissible uses remains appropriate
40:06 | and
40:06 | uh sufficient in accordance with changes
40:09 | in the laws.
40:10 | We've also checked on the vendor's
40:11 | contract terms, their data practices,
40:13 | system capabilities,
40:15 | and their proposed features that they've
40:17 | met
40:18 | um or that they require further review.
40:22 | We also look to see about additional
40:24 | privacy and any civil liberty laws,
40:27 | any technological issues, training,
40:29 | supervisor protections are needed based
40:31 | on the changes to the laws or other
40:33 | identified issues.
40:34 | We've also looked at the number of
40:37 | cameras we have, the placement, the
40:38 | deployment, and make sure that they
40:40 | relate to a
40:42 | public safety need.
40:43 | Um
40:45 | And we we assess whether the program
40:46 | should be continued, modified, reduced,
40:48 | or suspended.
40:49 | >> Is the police department the only
40:51 | units that have access to this
40:54 | information
40:55 | in terms of like it's not available to
40:59 | different town
41:00 | uh
41:01 | >> Like other town departments?
41:03 | >> No, just the Yes, other town
41:05 | departments.
41:06 | >> No, it's it's it's a
41:07 | It's Chelmsford Police Department owns
41:09 | the cameras and the data.
41:10 | >> Okay.
41:10 | >> Now, just to follow up with that,
41:13 | there's private individuals that could
41:14 | purchase a camera and they would get
41:16 | pictures of you know, if you're a
41:18 | business and you want to know who's
41:19 | coming in at what hours of the night. It
41:21 | I mean, there's nothing that prevents a
41:23 | private citizen from purchasing a camera
41:25 | privately.
41:26 | I mean, again, it's just going to show
41:27 | pictures and plates. They wouldn't have
41:29 | the ability to run the plate like the
41:30 | police would.
41:31 | >> But I can't speak for what private
41:33 | businesses do.
41:34 | >> And they don't have access
41:35 | >> 1,109 agencies as well do have access.
41:39 | The answer sounded like it's just
41:41 | Chelmsford, but there's also at least
41:43 | 1,109 agencies that have access. And and
41:46 | I can buy access to over 70
41:50 | police networks in Massachusetts for
41:52 | under $1,300.
41:56 | I would never do that, but that is under
41:58 | the radar, but absolutely
42:00 | >> I
42:01 | Yeah, so
42:02 | >> Why is Flock valuable to us in summary,
42:04 | okay? Faster investigative intelligence.
42:07 | It identifies vehicles quickly, turns
42:09 | the camera footage like a still image
42:12 | into something that we can work with.
42:14 | Um we've shown that it has proven public
42:16 | safety results. It's supported arrests,
42:18 | warrants, stolen vehicles, and the
42:20 | recovery of of at-risk juveniles and
42:23 | at-risk seniors.
42:25 | Better decisions, not just arrests. It
42:27 | helps us resolve misunderstandings and
42:29 | show when a vehicle was not involved.
42:32 | I'm going to bring up the bank robberies
42:33 | as a as an example, okay?
42:36 | We were given
42:38 | just multiple descriptions of cars, we
42:39 | were given multiple descriptions of
42:42 | plates, and what have you.
42:44 | It took us
42:45 | I was involved in it personally, it took
42:46 | us hundreds of hours
42:48 | to rule out junk, okay? When we could
42:51 | have been focusing on the the one person
42:54 | who was responsible.
42:56 | Um
42:57 | and
42:59 | some of these people
43:00 | had
43:02 | criminal histories related to robberies
43:04 | and related to other violent crimes, so
43:06 | you have to look at them.
43:07 | Um
43:08 | this system would have ruled them out,
43:10 | okay?
43:11 | Um and there are safeguards built into
43:13 | the program. We have a 30-day retention,
43:15 | we conduct audits, we limit sharing, we
43:18 | have data controls, and we review the
43:19 | system.
43:20 | Um
43:21 | that is it for my
43:23 | uh presentation.
43:25 | Um
43:26 | Karl uh Bock from Flock is on the Zoom
43:31 | right now. If you did have a question
43:33 | related to Flock corporate level,
43:36 | um
43:37 | you could address it to him, and then
43:39 | I'll stay and answer questions
43:41 | afterwards for whatever you need.
43:43 | So.
43:45 | Does anybody have a question for
43:47 | >> I did.
43:47 | >> Okay.
43:48 | >> a question for Karl.
43:49 | Um
43:51 | we talked about that
43:53 | the data is
43:55 | around for 30 days.
43:58 | My question is is it like
44:01 | January 1st to January 30th, and then on
44:04 | January 31st it gets dumped,
44:06 | or is it
44:08 | every day
44:10 | if something is 30 days or older, it
44:14 | rolls off?
44:16 | >> If it's 30 days or older, it rolls off
44:18 | is my understanding. Is that correct,
44:19 | Carl?
44:20 | >> That That's correct. It's a good
44:22 | question, and I think the best way to
44:23 | understand it is the moment a photo is
44:25 | taken, it's given an expiration date of
44:28 | when it should be deleted, and that
44:29 | happens automatically almost to the
44:31 | second to ensure that it's gone forever
44:34 | and it cannot be recovered.
44:37 | >> Carl, I have a question.
44:40 | Could you tell us, or do you have an
44:42 | idea how many accounts have been sold on
44:45 | the dark web, and perhaps what they've
44:47 | been used to access and and do?
44:49 | Can you give us any ideas on on the the
44:51 | downside, the negative sides of Flock as
44:53 | an entity?
44:55 | And and also, before I forget, um
44:59 | how many
45:00 | jurisdictions have I understand that
45:04 | privacy is not
45:06 | explicitly stated in the Fourth
45:09 | Amendment, but rights to a secure and
45:12 | you know,
45:13 | this How many federal agencies, how many
45:16 | federal
45:17 | have
45:18 | banned Flock for that reason, Fourth
45:20 | Amendment violations?
45:23 | >> Great. So, for the for the first
45:25 | question on cybersecurity and dark web,
45:27 | I'm not aware of access
45:30 | being able to be purchased through any
45:32 | kind of like nefarious actor. I but I
45:34 | will say this is a really serious and
45:36 | important cybersecurity question. We
45:38 | have mandated
45:40 | multi-factor authentication for each
45:42 | user in your department and all across
45:45 | the all across departments that use this
45:48 | to ensure that it's being accessed
45:49 | securely using best practices by
45:52 | by only authorized users.
45:54 | I will also say, you know, Lieutenant
45:57 | mentioned a lot of the cybersecurity
45:59 | audits.
46:00 | We just recently uh to conduct a
46:02 | third-party penetration test uh with a
46:04 | cybersecurity firm called Bishop Fox. Uh
46:07 | they were charged with being adversarial
46:09 | to try to penetrate looking for
46:11 | vulnerabilities in our software, in our
46:13 | hardware, in our cloud database, to
46:15 | identify any vulnerabilities so we can
46:17 | ensure that this information's being uh
46:20 | securely maintained. Um,
46:22 | so it's something that we need to be
46:23 | vigilant on as we know bad actors are
46:25 | going to be vigilant about that as well.
46:27 | Um, to your your second question about
46:30 | kind of the the legal landscape that
46:31 | we're in and how the courts have viewed
46:33 | the use of LPR um in Massachusetts but
46:36 | all across the country. You know, we've
46:38 | had more than 40 court cases uh that
46:40 | have shown that uh use of LPR
46:43 | information is consistent with the
46:45 | Fourth Amendment and in Massachusetts
46:47 | consistent with Article 14. Um, most
46:50 | recently in August uh there was a case
46:52 | um related to uh Commonwealth v. Saxon
46:56 | uh
46:56 | where a Superior Court judge ruled that
46:59 | um this the evidence gathered by LPR
47:02 | camera uh was uh allowed for the
47:05 | investigation of a of a of a homicide.
47:08 | Um,
47:08 | you know, as the courts continue to um
47:11 | you know, take these cases, you know,
47:12 | we're going to you know, make changes if
47:14 | we need to to stay consistent with state
47:16 | law. But um you know, as Lieutenant
47:18 | mentioned, the cameras are capturing uh
47:21 | an image in plain view literally looking
47:23 | for a government-issued document the
47:25 | license plate that has to be displayed
47:27 | on the rear of a vehicle um while
47:29 | traveling on a public road. Um, and
47:31 | because the there's limits to the time
47:34 | um
47:35 | and location. You know, these are fixed
47:36 | locations, the data is permanently
47:39 | deleted after the retention period, the
47:41 | courts have ruled that this is
47:42 | consistent with constitutional
47:43 | protections.
47:44 | >> Hey Carl, um if you're very familiar
47:47 | with all the
47:48 | uh cases that ruled them constitutional,
47:51 | I'm sure you're aware of the recent case
47:52 | in Oklahoma that ruled them
47:54 | unconstitutional under the Fourth
47:56 | Amendment at the federal level. Um
47:58 | wondering what your position is on that
48:01 | case as well. Um cuz they called out
48:04 | Flock specifically.
48:07 | >> Yeah, it's it's a more recent case that
48:08 | I'm not as familiar with. Um I will say
48:11 | that you know the the overwhelming
48:13 | jurisprudence on this at different
48:15 | levels of the courts have found this to
48:17 | be consistent with Fourth Amendment
48:18 | protections. And were that to change, we
48:21 | would we would make modifications to
48:23 | make sure that this is a tool that's um
48:25 | consistent with Fourth Amendment and uh
48:27 | protections from the Article 14.
48:30 | >> Could you could you tell us uh the four
48:32 | cameras we have and the two cameras that
48:34 | we have on order?
48:35 | Um I've seen several descriptions on how
48:37 | to hack it with the 3-second press of
48:39 | the button on the back of the camera.
48:41 | Are our cameras in town protected
48:43 | against that or are there any updates to
48:44 | the software that make it so anybody
48:47 | can't just go up there in then 5 seconds
48:49 | hack the cameras?
48:51 | >> Yeah, it's a good question and we're
48:53 | we're constantly pushing updates just
48:55 | like, you know, your your cell phone
48:56 | receives over-the-air updates to improve
48:59 | security
48:59 | >> Could I get a yes or no
49:00 | answer on that by any chance?
49:03 | Yes, our cameras are safe or or
49:06 | can our cameras be hacked with the
49:07 | 3-second button
49:08 | press?
49:09 | >> Yes, your camera your cameras are safe.
49:11 | >> our cameras are safe and cannot be
49:13 | hacked by a 3-second button press?
49:16 | >> Uh that's that's correct.
49:17 | >> Okay.
49:19 | Thank you.
49:20 | >> What does the green light do on so?
49:23 | Sorry.
49:24 | >> Uh how many Do you know how many uh
49:26 | currently installed roughly across the
49:28 | country? I've heard numbers from like
49:29 | 100,000 to 300,000.
49:32 | Uh and how many different agencies are
49:35 | there that that would have that access
49:37 | across the country?
49:39 | >> Yeah, I've I've seen I've seen a variety
49:41 | of numbers. I I think the number that
49:43 | was consistently reported is 120,000
49:46 | across 6,000 um communities across the
49:49 | country. We're operating in in 49
49:51 | states. Um but your your current
49:54 | deployment is the the four the four LPR
49:56 | cameras
49:58 | that are fixed to poles.
49:59 | >> How many clients do you have? How many
50:01 | different agencies are out there that
50:03 | would potentially
50:04 | >> Oh.
50:05 | >> Uh so
50:05 | >> That's not a question I get a lot. I
50:06 | know I know that for the law enforcement
50:08 | side it's it's 6,000 law enforcement
50:10 | agencies across the United States. Um
50:13 | having said that, you know, we mentioned
50:14 | that there's you know, a possibility for
50:16 | private contractors to to contract with
50:19 | Flock. So the the origin of the business
50:21 | is actually a homeowners association
50:23 | tool.
50:25 | Uh uh there's a
50:26 | series of break-ins in a suburb of
50:28 | Atlanta and the the cameras were created
50:31 | to give law enforcement a way to to
50:33 | investigate those crimes. Uh so
50:35 | homeowners associations are uh you know,
50:38 | entities we contract with, you know, big
50:39 | box retail as they um
50:42 | looked for evidence to investigate um
50:44 | organized retail theft. Um
50:47 | and as we said, you know, if they choose
50:49 | to, those private entities can share
50:51 | their network with local with their
50:52 | local law enforcement to aid in
50:54 | investigations. That's their choice. Uh
50:56 | but those those um those private
50:58 | contractors or private entities, uh they
51:00 | don't receive any of the hot list that
51:02 | we described earlier. So they wouldn't
51:03 | see an Amber Alert or a Silver Alert. Uh
51:06 | but they can
51:07 | grant permission to local law
51:08 | enforcement to conduct an investigation
51:10 | if there's, you know, a burglary or
51:13 | or an
51:14 | organized theft.
51:15 | >> But it all feeds into Flock, right?
51:17 | Everybody feeds into the major Flock
51:19 | databases.
51:21 | Obviously Flock has access to all to all
51:23 | of them. 6,000 or 10,000, how many
51:25 | clients you have. It all goes straight
51:27 | to you and then is
51:29 | you delete it in 30 days?
51:32 | >> So it's deleted from the the Amazon Web
51:33 | Service GovCloud automatically after 30
51:36 | days. And as they all sharing decisions
51:38 | are by the customer. So some some
51:41 | agencies choose not to share at all.
51:43 | Some share choose to share regionally,
51:45 | statewide, or nationally depending on
51:47 | their local policy and state law. That's
51:49 | something that's can be changed anytime
51:51 | by the customer and not by Flock.
51:54 | And we think that that local control to
51:56 | customize, you know, how you want to aid
51:58 | a neighboring jurisdiction investigation
52:01 | is really powerful.
52:02 | >> But but that that Lieutenant That data
52:04 | has been Flock has that data. Does Flock
52:07 | as an entity delete that data every 30
52:10 | days or does it have have hold on to it
52:11 | for over a year as stated by Flock
52:14 | itself?
52:16 | >> So so contractually it's set for a
52:19 | permanent deletion by the customer. So
52:21 | it's gone forever. Flock doesn't have
52:23 | access to it. It's gone.
52:25 | Jurisdictions can choose their retention
52:27 | level. So for example, New Jersey by
52:29 | state law has a 3-year retention period.
52:31 | Virginia by law, those photos have to be
52:35 | deleted after 21 days. So state law and
52:38 | local practice can dictate that as well
52:40 | as local policy.
52:41 | >> But aren't they on a cloud server?
52:44 | >> On the On the Amazon Web Service Cloud,
52:46 | yes. And that's where they're scheduled
52:48 | for permanent deletion after the
52:50 | retention period expires.
52:52 | The The only way to retain it would be
52:54 | if it's downloaded for evidence in a
52:57 | trial, and then it would follow and then
52:59 | it comes off the Flock, you know, the
53:01 | the the PDF Lieutenant showed is that
53:04 | that download. That would follow the
53:06 | regular courts evidentiary retention and
53:08 | then be eligible for discovery or
53:10 | whatever is required for a court.
53:13 | >> A quick question to the Lieutenant.
53:15 | >> I'm going to If Does anybody else have
53:17 | questions for Carl?
53:18 | >> I I have a question. Yes, so my question
53:21 | is has Flock ever been hacked?
53:26 | >> It's a good It's a good question. So our
53:28 | our our Amazon Cloud
53:30 | system has never been breached. There
53:33 | have been limited instances where an
53:35 | individual camera was accessed and some
53:37 | of the data that's temporarily stored
53:40 | before it's sent to the cloud was
53:41 | accessed,
53:43 | but that's a limited um limited example.
53:46 | It it literally could be you know the
53:48 | same thing you or I could witness at a
53:49 | busy intersection as we're taking
53:50 | photos.
53:52 | That breach, you know, has been you know
53:54 | identified and mitigated, but at no
53:57 | point was the larger cloud
53:59 | infrastructure or
54:01 | what was was breached or hacked. It's
54:03 | something that we are constantly trying
54:05 | to stay vigilant on. We recently
54:08 | announced a bug bounty program for
54:11 | ethical hackers. If they were to
54:13 | identify a vulnerability, they can
54:14 | report it to us and get credit for for
54:17 | finding a flaw.
54:18 | We think that this is something that's
54:20 | really important that we need to stay
54:21 | up on.
54:22 | >> Is it true that a hacker was the one who
54:24 | actually pointed out the vulnerabilities
54:27 | to you?
54:29 | >> In some in some
54:31 | some of those cases
54:31 | >> blog post that was brought to federal
54:33 | court. Is that true? Are you
54:37 | >> You guys are full of
54:39 | >> Just
54:40 | Yeah.
54:41 | >> So, I I like to follow up on the
54:43 | >> Maybe maybe
54:44 | >> I'd like to follow up on the
54:45 | >> closing I can just say
54:47 | >> But
54:47 | >> Yeah.
54:48 | >> I'd like to follow up on
54:49 | >> Maybe in closing I I can just say that
54:51 | the the couple of examples I'll send you
54:54 | >> Carl, there's one gentleman in the back
54:55 | that's been trying to ask a question.
54:56 | >> Oh, I'm sorry. Go ahead.
54:58 | >> So, just to follow up
54:59 | Yeah, I'm under the impression that
55:01 | Amazon's
55:03 | cloud services are pretty secure.
55:06 | But there was an article
55:08 | about Ledyard, Connecticut. You know,
55:10 | there's a casino in Ledyard. Maybe
55:12 | you've heard of it before.
55:14 | So, actually there are multiple articles
55:16 | about this. So, it's not just one news
55:18 | source.
55:19 | Uh basically it was Flock warns
55:21 | Connecticut police about security
55:23 | breach.
55:26 | Um and the records involved camera
55:29 | locations, camera
55:31 | and and other Flock devices, you know,
55:33 | like the type of device whether it's
55:35 | camera, Raven, microphone, whatever.
55:39 | And it's on a national scale, but for
55:40 | some reason this article said they
55:42 | warned Connecticut police. I'm not sure
55:43 | if they warned Massachusetts police,
55:45 | too, but
55:46 | this was apparently a big deal down the
55:48 | road in Ledyard.
55:50 | And they canceled their contract. Um
55:54 | I've seen
55:55 | the website that they referred to in
55:57 | this article, and it's on a national
55:59 | scale.
56:00 | You can see all the cameras in the
56:01 | entire country.
56:03 | And a bunch of those cameras
56:05 | >> It was referred to as a security breach.
56:06 | So, data breach, security breach,
56:08 | whatever.
56:09 | I don't know if it had to do with the
56:10 | Amazon cloud, but it was called a
56:12 | security breach.
56:15 | >> Yeah, that there's an there's an ongoing
56:17 | investigation, law enforcement
56:18 | investigation on that. Um I will say
56:20 | that the uh
56:22 | the Amazon Web Services Cloud uh was not
56:26 | accessed or breached. Um camera
56:28 | locations are something that we talked
56:30 | about tonight. These are devices that
56:32 | are in plain view. Um a lot of crowd
56:34 | sourcing has identified the locations of
56:37 | these these devices in the in the
56:39 | community. So, um
56:40 | >> I've heard that they're trying to make
56:41 | >> to answer your question, the the the
56:43 | Amazon uh Web Services
56:44 | >> to make it illegal to have something
56:46 | trying to trying to make it illegal to
56:47 | deploy
56:47 | >> to that. Just
56:48 | let
56:49 | >> Let Let the guy finish what
56:50 | >> Just let him finish answering his
56:51 | question.
56:52 | >> All right.
56:53 | >> Correct. Yeah, the the the cloud
56:54 | database was not breached.
56:59 | >> I have a question regarding the actual
57:00 | hardware.
57:02 | Uh we have Falcon. Is that the model we
57:05 | have?
57:05 | >> That's what we have, Falcon.
57:06 | >> Okay. Yes. Okay. Now, our Falcons we
57:09 | we're using them as just a
57:12 | plate reader.
57:13 | Is there any chance that
57:16 | the capabilities of that Falcon camera
57:19 | can be changed or upgraded to do facial
57:23 | voice recognition or
57:25 | is that hardware set? If there's a
57:27 | software update, can does it have more
57:29 | capabilities?
57:32 | >> So, the to answer your question, the we
57:35 | sell a single-purpose tool. That's the
57:37 | the LPR camera that takes the still
57:39 | video still photo of the rear exterior
57:42 | of the vehicle.
57:44 | Um it does, as we discussed, uh identify
57:47 | the state and license plate um
57:49 | information from the vehicle. It also
57:51 | looks for key vehicle characteristics,
57:53 | make, model, and color. So, that's kind
57:55 | of the enhancement of that. Um and I
57:58 | think the reason that's important and
57:59 | was mentioned a couple of times with the
58:01 | with a hit-and-run investigation, for
58:02 | example, the you know, the pedestrian
58:04 | struck by a vehicle, um it's unlikely
58:07 | that they'll have the confidence to
58:09 | remember the exact license plate, but
58:11 | with make, model, and color, they can
58:13 | give that to a detective to aid in the
58:15 | investigation to determine, you know,
58:16 | that you know, a suspect vehicle
58:18 | associated with that crime. So, it it is
58:20 | more than a just a license plate reader.
58:22 | It does look at the uh the exterior of
58:24 | the vehicle. Uh that can provide
58:26 | actionable investigative lead for for
58:29 | you know, justice
58:30 | >> for victims and to
58:31 | to help investigations.
58:32 | >> I understand that, but my question was,
58:35 | is there a possibility of increased
58:38 | um capabilities of the actual cameras
58:40 | that we own upon software upgrade?
58:46 | >> I I I can say for facial recognition,
58:47 | there's no way. Uh that's not something
58:49 | that we're interested in pursuing. Our
58:51 | customers aren't asking for it. Um at
58:53 | the point there is any kind of like
58:55 | additional software or additional
58:56 | devices that your you know, jurisdiction
58:59 | wishes to purchase, that would be an
59:01 | additional purchasing agreement that you
59:03 | would have to enter into.
59:05 | >> And they don't have facial recognition
59:07 | or voice recognition capability,
59:09 | correct?
59:10 | >> Correct. Yes.
59:11 | >> a software update.
59:12 | >> Even the hardware does not allow that,
59:14 | correct?
59:16 | >> We We do not do facial recognition. Uh
59:19 | the cameras are cameras. They wouldn't
59:20 | be able to do
59:22 | voice recognition.
59:23 | >> No, no. We know that it's It's doing it.
59:26 | Is it capable of doing that? Is it
59:29 | capable of upgrading with software
59:31 | upgrade?
59:32 | >> There's other companies offering to
59:34 | integrate all the information, so Flock
59:35 | keeps their hands clean and say they
59:37 | doesn't do facial recognition and they
59:39 | probably never will, but the Dizzimo
59:41 | promises that they can and they're
59:42 | actively getting people they're actively
59:44 | recruiting different police
59:47 | organizations to say, "Yeah, we want
59:48 | it." and then they'll finish developing
59:50 | it and sell it. So, what's up with all
59:52 | these Flock license plate readers in
59:54 | parks and playgrounds and places where
59:57 | there aren't license plates?
59:58 | >> Carl.
59:58 | >> What what what is the what are those?
60:00 | >> I just want to make sure you have a
60:01 | question. Carl, the Falcon does not have
60:04 | the capability
60:05 | the hardware does not have the
60:07 | capability for voice or facial
60:09 | recognition, correct?
60:10 | >> Ever.
60:10 | >> Correct.
60:11 | >> But the photos could be used by another
60:13 | company to accomplish the exact same
60:15 | thing.
60:15 | >> but not this.
60:17 | >> Carl, the face.
60:18 | >> Hold on one second. We have one more.
60:19 | So, do the Falcon model cameras can that
60:23 | feed into the Flock free form software
60:26 | that allows like a search by clothing?
60:29 | Like Flock's own website has like this
60:32 | example of searching for a man in a camo
60:33 | hat and a forest green sweater. Can the
60:36 | Falcon model cameras provide detail like
60:38 | that? The images look very clear
60:41 | contrary to a previous meeting that
60:43 | we've had where
60:45 | either the select board or officers here
60:47 | said they were not very clear photos,
60:49 | but they're very clear. So, do those
60:50 | cameras feed into that Flock software
60:53 | system and allow for this type of
60:54 | search?
60:57 | >> So, for LPR and for the Falcons, you
60:59 | cannot search for any
61:02 | person or person characteristics.
61:05 | We do sell like live video camera and we
61:08 | do offer software upgrades for for for
61:11 | live video.
61:12 | That's that's an option that you've
61:14 | chosen not to take, but no, the the
61:16 | Falcon cannot be used to search for
61:18 | human people characteristics. It's
61:20 | literally like plastic and metal.
61:22 | The exterior of the vehicle and
61:24 | attributes related to that.
61:26 | >> How long do these cameras last? Like
61:28 | when will they need to be replaced?
61:30 | We've already had them installed for a
61:31 | couple years. We're assuming at some
61:33 | point when there's better technology we
61:34 | might want to upgrade. I mean I would
61:36 | upgrade.
61:38 | >> So we upgrade on a lease subscription
61:40 | model.
61:41 | So it's a fully inclusive lease. Each
61:43 | camera is $3,000 for a year. That
61:45 | includes all the cloud storage,
61:48 | number of authorized users,
61:50 | and and periodic upgrades. So at the end
61:52 | of your contract if you choose to renew,
61:55 | you we'd want to make sure that you have
61:56 | the most up-to-date device on the pole.
62:00 | You know, I talked to you know, police
62:02 | chiefs all the time and they say there's
62:04 | there's a room in every police station
62:06 | where there's just all the old
62:07 | antiquated printers and technology
62:09 | that's out of step. So by you
62:11 | contracting with us with a lease
62:12 | subscription, you we're making sure that
62:15 | the device on the pole is up-to-date and
62:17 | accurate.
62:18 | >> So you So So part of our $3,000 a year
62:22 | gets an upgrade when it's available or
62:23 | when the camera runs out? Like how soon
62:26 | do we have to replace them?
62:28 | But I I missed I didn't
62:29 | >> Yeah, I mean that like
62:30 | >> We're going to replace
62:31 | >> That that would replace them. Or we're
62:32 | going to run them until they die and
62:33 | then replace them once they're
62:35 | >> Yeah, that that would come up in your
62:36 | renewal conversation to determine the
62:38 | state of the device and what's available
62:40 | to make sure that you're you know,
62:43 | using the best technology available.
62:45 | But
62:46 | maintenance and
62:48 | upkeep that's all included with the
62:51 | with the subscription model.
62:55 | >> Does anybody else have any questions for
62:56 | One more in the back.
62:58 | >> Yeah, so I have a question. Would you be
63:00 | able to tell us about the I think it's a
63:02 | software product called OS Investigate
63:06 | by Flock?
63:10 | >> Yeah, so we we have a number of hardware
63:12 | and software device
63:13 | packages. This is a tool that you've
63:16 | chosen not to pursue.
63:18 | Um,
63:19 | just like the live video camera or drone
63:21 | or the other kind of software
63:23 | enhancements. Um, the OS investigate is
63:26 | a additional
63:28 | uh software purchase that um brings
63:30 | together different amounts of
63:32 | information that's already available uh
63:34 | to local law enforcement to help with
63:36 | the investigation.
63:37 | >> Can you be a little more specific about
63:39 | that? Like give an example of its use.
63:46 | >> Oh, yeah. It's not something that I'm
63:48 | really well versed in. I'm happy to
63:50 | follow up or provide more information on
63:52 | our website about the capability. Um,
63:54 | you know, we were looking at uh
63:55 | presenting on the technology that you
63:57 | have tonight in your community and and
64:00 | not something that you decided not to
64:01 | purchase.
64:03 | But I there's information on the
64:04 | website, I'm sure.
64:06 | >> Just one small thing. Uh Pearl
64:10 | the cloud storage, because, you know,
64:13 | the data belongs to Charlton only.
64:16 | Uh
64:18 | how frequently are backups done? I mean,
64:20 | is it like a daily backup of everything?
64:22 | And as are the backups of our data
64:25 | separate from anybody else?
64:28 | And how many backups occur?
64:32 | >> I don't I'm not aware of any backup. It
64:34 | I mean, the the system is designed so
64:36 | that when the retention period expires,
64:38 | that data in its entirety is permanently
64:41 | and irrevocably gone. So, I
64:45 | That I mean, that's that's the way the
64:46 | system is designed so that you know,
64:48 | when when we we say it's gone off the
64:50 | Amazon web service cloud, it's gone
64:52 | forever.
64:54 | >> But I have a question. But what happens
64:56 | I have a question.
64:56 | >> We have one last question in the corner
64:58 | here.
64:59 | >> How does it work though if you every 30
65:01 | days this data is
65:04 | disappears, okay?
65:06 | But then the crime happens and then your
65:09 | court system or your lawyers need to
65:11 | find what happened on this day what 3
65:14 | years ago?
65:15 | How does Flock help you with the
65:16 | investigation if all that data is
65:19 | deleted?
65:20 | >> So, if
65:21 | >> Does that make sense?
65:22 | >> On our end?
65:23 | >> Yeah.
65:23 | >> That's Carl touched on a minute ago. If
65:26 | we had a specific vehicle, we download
65:29 | that that instance for that vehicle to
65:32 | preserve for evidence.
65:35 | But, if we were looking at if you were
65:38 | come home from Florida
65:40 | and you report, "Hey, my house was
65:41 | broken into." And you realize it's 60
65:43 | days ago. We're kind of out of luck
65:45 | because that 30-day window and we've
65:48 | tried to
65:49 | we've tried to get beyond it to see if
65:52 | we could and we can't. Like, we've had
65:55 | no success in doing that.
65:57 | Yes.
65:58 | >> And that's So, it helps if we know
66:00 | there's a crime that occurred cuz then
66:01 | we can save the data.
66:03 | But, if we don't know a crime, then we
66:05 | wait 15 45 days and we
66:07 | miss that opportunity and that's the
66:09 | problem. With something that's not
66:11 | immediately reported. But, if there was,
66:13 | you know, what to Anthony's point, we
66:14 | could retain something if we knew it was
66:16 | a stolen car, we could say, "Hey, this
66:18 | went by here. We don't know where it
66:20 | went, but it went by at 2:00 in the
66:22 | afternoon." We could save that because
66:24 | we know there's a crime of a stolen car.
66:26 | But, we couldn't do it 6 months from
66:28 | now.
66:28 | >> Right.
66:29 | >> That data would be It could have driven
66:30 | by at 2:00, but we won't know.
66:33 | >> Yeah, that's where I thought it would
66:34 | be. I know it's a helpful tool for you
66:36 | guys, but then I'm thinking, "Well, if
66:38 | it gets deleted every 30 days and
66:41 | something happens 3 years ago, how do
66:43 | you
66:43 | >> Yeah, it's not going to help us.
66:45 | It's not crystal ball. It's not going to
66:47 | help us 3 years ago. It's not. It's But,
66:49 | if we know about a crime right now, we
66:51 | could certainly save that data.
66:53 | >> Right.
66:54 | >> We can download that data and save that
66:56 | data.
66:57 | Um like we would with anything else
66:59 | like, so. Um you know, somebody could
67:02 | have a warrant today,
67:05 | clear it tomorrow, the case still goes
67:07 | to court, and we retain a copy of the
67:09 | warrant.
67:09 | >> You see that the officers wear body worn
67:11 | cameras?
67:13 | I'm not wearing one, so I can't point to
67:14 | it. He said not wearing one right at the
67:15 | moment, but you'll see the uniformed
67:17 | officers are usually wearing a body worn
67:19 | camera.
67:20 | They have to record in each interaction,
67:22 | and there's a time frame like if hey,
67:24 | this was just a
67:25 | uh a nonsense thing, we had a
67:26 | conversational, that might get deleted,
67:28 | you know, random yeah or something.
67:31 | However, because is there's only so much
67:32 | storage in the cloud that we can
67:34 | utilize.
67:35 | But if somebody said, "Oh my god, this
67:37 | was a you know, this guy hit so-and-so
67:39 | with a bat." They watched it, they would
67:40 | definitely save that, they'd label it so
67:43 | that it would retain longer because
67:45 | that's an actual crime and not a simple
67:47 | conversation.
67:49 | Does anybody else have any questions for
67:50 | Carl, or are we good? Thanks, Carl.
67:52 | >> I never got a description of what
67:53 | happens at the playground and why you
67:54 | got cameras there.
67:55 | >> I'm sorry, say that again?
67:56 | >> I never found out why there's flock
67:57 | cameras at playgrounds when there's not
67:59 | >> Oh, well, that depends.
68:00 | >> I I couldn't answer that. Carl, could
68:02 | you?
68:04 | >> Is it Is this like a camera you deploy
68:06 | to a
68:07 | playground or is it
68:08 | >> we only have four public ways. We only
68:10 | have four, there's 120 or 300,000 if you
68:13 | want to be honest, installed already
68:15 | across the country. So, one of those
68:17 | other ones.
68:18 | >> Are there not any installed at It
68:19 | weren't you giving a demo Somebody was
68:21 | giving demonstrations at a gym? It was
68:23 | something like that. I I've
68:25 | I've seen that story a hundred times.
68:27 | >> Yeah. So So to the again, my earlier
68:30 | point, we sell a number of different
68:31 | devices. Some include live video
68:33 | cameras.
68:34 | In some jurisdictions where they see
68:36 | vandalism in public parks after hours,
68:39 | they've found that a live video camera
68:42 | is helpful to invade to aid in an
68:44 | investigation or to alert
68:46 | to to alert local law enforcement if
68:50 | you know, if there's a trespassing or
68:52 | you know, other other crimes happening.
68:54 | Again, that that's not a technology that
68:56 | that you chose to deploy, but that that
68:58 | would be a use case.
68:59 | >> Those do like jackets or something.
69:01 | >> didn't buy any of those. So, they're not
69:03 | here. He's telling us about what we have
69:05 | here.
69:06 | >> Yeah, but I
69:07 | think we could put in anything else any
69:08 | other time and in a couple years we're
69:10 | going to have to adjust it and nothing's
69:12 | stopping us from buying four more after
69:14 | the four we have on the way. So, I want
69:16 | to know what they're about. Just because
69:18 | we don't have them in town here doesn't
69:19 | mean I'm not driving in Worcester
69:20 | tomorrow or or any number of other
69:22 | places in this country. Uh just because
69:25 | Charlton didn't buy them doesn't mean
69:26 | they're not good.
69:27 | >> this gentleman is telling us about
69:30 | Flock, what Charlton, not Flock.
69:32 | >> Well, well, he's he's from Flock. He can
69:34 | tell us about Flock all over, right?
69:36 | >> It's a general question of Flock.
69:38 | >> All right, it's not I mean, it's not our
69:40 | four current cameras and probably not
69:41 | our four next ones, but we need to know
69:44 | about about as much as might be
69:46 | installed without us really knowing
69:48 | about it cuz we approved it or forgot it
69:50 | to go to the meeting to not approve it
69:51 | like two years ago. Now they can put in
69:53 | whatever they want. So, I think it's a
69:54 | fair question.
69:56 | Just what's in Charlton? I don't live
69:58 | only in Charlton. I I I drive around.
70:00 | >> I can't speak for another town or
70:01 | another community.
70:02 | >> to Flock, right? I don't expect to
70:04 | what their technology are, what's it
70:05 | doing, what like
70:08 | and and what that really does for you it
70:10 | recognizes
70:12 | body shape and all that other stuff. It
70:13 | might not have facial recognition now
70:15 | like you said, but it does have other
70:17 | uses. So, I was just trying to get more
70:18 | clarity on that, but I guess we'll stick
70:20 | to Charlton for now.
70:22 | >> All right, everybody's
70:23 | Thank you, Carl.
70:25 | >> Thank you.
70:26 | >> Thank you. I appreciate it.
70:33 | >> Okay, does anybody have any specific
70:35 | questions to Charlton that we can help
70:36 | you with?
70:38 | >> No.
70:38 | >> Okay.
70:39 | >> What Where are we going to put those
70:41 | other four cameras if we do?
70:43 | >> Um we've put everything on hold so we
70:45 | don't
70:46 | >> And what do we do if I know this is a
70:49 | non-binding thing coming up in a week or
70:50 | whatever. What do we do if you get uh
70:52 | strong resistance to doing Flock? We're
70:54 | going to talk about this? We're going to
70:56 | take them down? We're going to at least
70:57 | stop
70:58 | putting up the form more if everybody
71:00 | votes to take them down. Can we at least
71:01 | stop putting up more until we
71:04 | Like where do we go? Where do we come
71:05 | back to that? If the vote is
71:06 | overwhelmingly people
71:07 | >> Thank you. No problem. Thank you.
71:08 | >> Sorry.
71:09 | >> The select board can override the vote
71:12 | next week or whatever.
71:14 | >> Well, it's not binding. So, the select
71:15 | board has to actively choose to take it
71:17 | up if they want to.
71:18 | >> But if they override it, we can move
71:20 | forward to a binding, right?
71:22 | >> Yeah, we have We have a binding thing in
71:23 | with the ACLU now.
71:25 | >> to do it in a in a different way.
71:27 | >> Okay.
71:28 | >> James, you
71:31 | >> The places I'm interested in putting
71:33 | them, if we get the you know, the four
71:35 | cameras go up, would be Route 20 mainly
71:37 | cuz we've got a lot of people
71:39 | >> Thank you.
71:40 | >> Well, they A lot of folks are coming out
71:41 | of Connecticut stealing cars. We just We
71:43 | just And we know they're going back and
71:45 | forth. That's kind of If we knew each
71:48 | end, we'd know at least they were in
71:49 | town, out of town. You know, we'd know
71:51 | that if we had one on each end of 20.
71:53 | >> I'm not here to deny that there are some
71:54 | good positive uses with this camera,
71:56 | right? I mean, the large majority of it,
71:58 | that's what's why it's tough to get
71:59 | anybody uh to see the the bad part of
72:02 | it. I mean, it's It helps It helps the
72:04 | police.
72:05 | >> It's all, you know, nobody wants to be
72:06 | followed. I mean,
72:08 | it's
72:08 | >> about this multiple times. Like I have
72:10 | an anti-police sentiment and sentiment.
72:12 | And that's not it. I I'm kind of more uh
72:15 | freedom and lack of surveillance and all
72:17 | that stuff. I mean, where it can go.
72:19 | Where it is today might seem safe,
72:20 | especially in this nice small town with
72:22 | good officers,
72:24 | you know, and stuff like that. You know,
72:26 | having but it
72:28 | feeds into everything else. It's It And
72:31 | it can get so much worse. Just because
72:33 | my phone always tracks me, doesn't mean
72:35 | I want something else to track me. Just
72:37 | because you've been stabbed once doesn't
72:38 | mean you want You want to be stabbed
72:40 | again.
72:40 | >> wrong, but nobody's willing to go back
72:42 | to landlines, either. Like the
72:44 | technology's there, and when you drive
72:46 | down the street, whether you want to
72:47 | admit it or not, you're in public. If I
72:49 | decide to wear an offensive shirt and I
72:51 | go walking down the street, that's a
72:53 | little different than if I wear that
72:54 | same offensive shirt in my house, and
72:55 | people are sneaking around and looking
72:57 | in my window. Well, that would be No,
72:59 | that would be wrong. You've got a right
73:01 | to privacy in your home. But, when you
73:03 | go out in public, and I I wear something
73:05 | offensive, and I walk down the street,
73:07 | it's fair game. That's the same thing
73:09 | with a license plate. You don't have You
73:11 | have to have a license plate on your
73:12 | vehicle. You have to have it registered
73:14 | and insured when you drive down the
73:15 | road. Those are rules that and laws
73:17 | we've all agreed to over the years. But,
73:19 | just
73:20 | >> don't have to wait let the state conduct
73:22 | like a search on you and
73:25 | create a database against you. And the
73:27 | state doing it is different than Google
73:29 | doing it. Just cuz I need their stuff
73:30 | and I click on yes, I'll do whatever you
73:32 | say. Nobody's reading that, whatever it
73:34 | is. Just because I say yeah, Google can
73:36 | look at my pictures, whatever, which I
73:37 | don't want them to, doesn't mean I want
73:39 | the state doing it. I mean, hey, fine.
73:41 | We You know, like if we can sit down and
73:42 | have coffee together, but I don't need
73:43 | you to know everything about me. Uh and
73:46 | and somebody else somebody with a bad
73:47 | actor to do that. That's what I'm trying
73:49 | to stop. I mean, where we're at now, if
73:50 | we're just doing it to to catch
73:52 | speeders, or we're not even doing
73:54 | speeders. We could not
73:55 | >> We're not doing speeders. We're not
73:56 | doing it.
73:56 | >> But, this all could be integrated in.
73:58 | >> for you, too.
73:59 | >> Okay.
73:59 | Okay.
74:00 | >> It's just it's it's
74:01 | um it's just your thoughts.
74:03 | Um
74:04 | now, from what I read in an article
74:07 | um in the Worcester Telegram and
74:08 | Gazette, I read that we are the only
74:11 | town in Worcester County left with
74:14 | functioning Flock cameras. Is that
74:16 | correct?
74:16 | >> I don't know.
74:17 | >> I don't know, to be honest with you.
74:18 | >> Okay, so that was what was reported by
74:20 | the investigative reporter.
74:22 | But,
74:23 | I actually have the article. I can send
74:24 | it
74:24 | >> No, no, no. Can you send me an I gave
74:26 | you my email earlier. Can you send me an
74:28 | email, and I will look for you and find
74:31 | it?
74:31 | >> Sure.
74:31 | >> Because
74:33 | I see Flock
74:34 | cameras in other towns, and
74:37 | >> But,
74:37 | >> we put them up.
74:38 | >> They have they have been
74:40 | decommissioned. They're not
74:41 | >> Somehow, but I don't think so.
74:42 | >> Okay, so that was my first question.
74:44 | Then my second question to you is, what
74:46 | are your feelings and your feelings on
74:49 | the large amount of towns surrounding us
74:53 | that have abandoned the technology? What
74:55 | what are your feelings on that? Do you
74:57 | think that we stand out, that we need it
74:59 | more than everywhere else? Like what
75:01 | what is what is it?
75:01 | >> I would say this,
75:02 | from my perspective, we were one of the
75:05 | first people to
75:07 | let's say jump on board with this stuff.
75:09 | We saw a lot of success.
75:11 | Um
75:12 | and a lot of the towns that have jumped
75:14 | off of Flock
75:16 | have inquired of us for Flock data,
75:19 | okay?
75:20 | >> using you guys?
75:21 | >> No, so they've inquired We
75:23 | We're selective in in what we do, but
75:26 | I'm saying prior to this, okay? They
75:27 | used us like
75:29 | >> So before they shut down?
75:31 | >> South Rich had a shooting, they would
75:32 | call us. Worcester, you know, other
75:34 | places have a shooting
75:36 | or another crime, they'd call us for
75:38 | Flock data, okay? I can't tell you why
75:41 | other towns
75:43 | >> Well, we know why.
75:43 | There's many of the towns that have
75:45 | released public statements released why
75:47 | they've abandoned the Flock technology.
75:49 | >> there's a campaign. I'm aware of that. I
75:51 | realize it's an organized campaign to to
75:53 | get rid of cameras.
75:54 | >> Worcester did not organize
75:56 | >> But well,
75:57 | you're here at a certain time and place,
75:59 | so I you're organized somewhat.
76:02 | And that's fine. Everybody's entitled to
76:03 | their own opinion. But
76:05 | a lot of times things get grouped as
76:07 | Flock. There's other products out there
76:09 | in other communities.
76:10 | >> We'll say Flock and Flock-like products.
76:12 | >> That's all. You know, I'm just saying.
76:13 | So whether it's Axon or Motorola or you
76:17 | know, I I got to go
76:18 | on a trip to the airport and I'm going
76:20 | to, you know, fly. I mean, I'm
76:21 | unfortunately I'm going to go walk
76:23 | through the airport and they're going to
76:24 | x-ray me and they're going to take check
76:26 | my bag and they're going to probably
76:28 | facial recognition that I'm unaware of
76:31 | and I
76:32 | it's just Does it make me feel like this
76:34 | things are great? No, it don't. It's a
76:36 | little eerie, but at the same time after
76:39 | 9/11 nobody had a problem taking off
76:41 | their shoes and going through a metal
76:43 | detector because it was like a, you
76:44 | know, a serious thing. I don't want the
76:46 | average person followed around or have
76:48 | any thoughts that we're trying to
76:50 | surveil them or
76:52 | >> I had never felt
76:53 | >> There's hundreds and hundreds of
76:54 | thousands, I would Anthony would correct
76:55 | me, of pictures that just disappear.
76:57 | Like they're not even We don't even look
76:59 | at There's days we go by without looking
77:01 | at Flock. But a department will call us
77:03 | and say, "We just had a hit-and-run."
77:06 | It's a lot easier if somebody said,
77:08 | "Hey, I just got a a red pickup truck
77:11 | ramming off the road went by the
77:12 | library." It's a lot easier to go, "When
77:13 | did it happen?" Okay, well, only four
77:15 | pickup red pickup trucks went by in the
77:17 | last hour. We only have to talk to four
77:20 | people versus
77:22 | there's 4,000 red pickup trucks in the
77:24 | area. We'd have a lot of suspects. So,
77:25 | that's that's where it helps us.
77:27 | >> I really appreciate the fact that you
77:30 | guys are We have an amazing police
77:33 | department. I really appreciate your
77:36 | passion and efforts for all this. And
77:38 | I'm I thank you for your answers.
77:40 | >> Um again, your time
77:41 | >> And your time tonight, yes. And um so so
77:44 | if I'm hearing you correctly, um when I
77:46 | asked you how you felt about the
77:48 | surrounding towns, you're looking at us
77:50 | as independent an island. Like we this
77:53 | is good for us. This is what we need. Is
77:54 | that what I'm hearing?
77:56 | >> Yeah, I I I think as as police chief, I
77:58 | know
77:59 | >> And that's where you're coming from that
78:00 | that's the justification for
78:01 | >> I know that things could have been
78:03 | different if these cameras were up in
78:05 | something like Holly Perini
78:07 | disappearance or Molly Bish. Like we
78:10 | might have had a suspect
78:12 | like that.
78:13 | >> Okay.
78:13 | >> You know.
78:14 | >> That's
78:14 | >> Do I want to
78:16 | you know, anybody I don't I'm not
78:17 | interested in I don't want people
78:18 | speeding going through red lights, but
78:20 | at the same time I'm not interested in
78:21 | tracking that down. I'm zero interest.
78:23 | >> Okay, that's what
78:24 | >> Zero interest. That's what police
78:25 | officers should go out there and
78:26 | unfortunately, if you decide to go
78:28 | through the red light and you get pulled
78:29 | over, that's on you. I don't make I
78:31 | don't think we need Big Brother
78:33 | Ooh, you know, you went 3 miles an hour
78:35 | over the speed limit. You're you a
78:37 | ticket in the mail. I don't I don't
78:39 | That's not what I'm looking for. I'm
78:40 | looking for serious crimes and you know,
78:44 | obviously there's minor crimes too, but
78:45 | you know, property damage when you're
78:47 | upset when somebody vandalizes your car
78:49 | or breaks into your car and you want us
78:51 | to and all we can say I don't know,
78:53 | nobody was there. Well, we can look at
78:55 | it well, all right, if it's 3:00 in the
78:56 | morning, and they probably drove down
78:58 | this street, there's a good possibility
79:00 | that there's only 10 cars. We got a What
79:02 | kind of a color was it?
79:03 | >> I'll give you one example like in line
79:06 | with what the chief said, okay?
79:08 | We had somebody try to falsely report a
79:10 | crash. Uh
79:12 | um
79:13 | Again, I'm going off my memory, so
79:14 | don't, you know, quote me on this. And
79:17 | this goes back a year, year and a half.
79:19 | Somebody tried to falsely file an
79:21 | insurance claim saying that their car
79:23 | was hit.
79:24 | And there was a flock image of them
79:26 | driving by
79:28 | with the damage on and then the flock
79:31 | camera with the same damage. They tried
79:33 | to falsely report it.
79:35 | They came to the police station, go to
79:37 | file a report saying that their car was
79:40 | involved in a hit-and-run.
79:42 | And we were we were able to say, "Hey,
79:43 | no. Like, you had this damage ahead of
79:45 | time. You're trying to like falsely
79:48 | say something that didn't happen, okay?"
79:50 | The same as we've excluded people. We've
79:52 | excluded suspects in crimes. We've
79:55 | uh
79:56 | we've used it to solve crimes, but we've
79:58 | also used it to exclude people out of
79:59 | crimes, too.
80:01 | And the biggest thing is, in my opinion,
80:04 | the biggest benefit is
80:06 | an elderly person, suicidal kid,
80:08 | suicidal adult,
80:10 | a person goes missing, a person is
80:11 | kidnapped. The The value of that, in my
80:14 | opinion, in
80:15 | returning somebody to their house, you
80:17 | can't put a dollar sign on it. Um you
80:19 | know,
80:20 | our property damage
80:22 | crimes that we're able to solve and
80:24 | property loss,
80:25 | um at least somebody gets restitution.
80:27 | Somebody gets it, you know, somewhat of
80:30 | being made whole. But those crimes where
80:32 | somebody might kill themselves or uh
80:35 | you know, somebody wanders off and they
80:37 | you know, they drive into a down a dirt
80:39 | road, those we can't put a dollar sign
80:41 | on those. And those to in my opinion,
80:44 | that is where the value is.
80:45 | >> So, it sounds like both of you are
80:47 | pretty kind of scared of like the more
80:50 | advanced capabilities of this system and
80:52 | wouldn't want that in our town, but
80:54 | >> We don't have a need for it.
80:55 | >> Yeah. And but say like,
80:57 | >> I don't know.
80:58 | >> 15 years down the line, both of you are
81:00 | not in the department or something like
81:01 | that, how do we ensure that the next
81:03 | person in your place has a similar
81:06 | feeling about like because the system is
81:08 | fully capable of like person searches
81:10 | and video things like that, how do we as
81:13 | a town ensure that we keep it at a level
81:15 | that both of you are saying is very
81:17 | beneficial to our department um and is
81:19 | not super invasive, I guess.
81:22 | >> I I would say that I mean, I maintain
81:25 | it, you maintain it, and I don't think
81:27 | to be honest with you, if you surveyed
81:28 | the police department, I mean,
81:29 | I I don't think there's anybody
81:30 | interested in in those capabilities
81:32 | because
81:33 | >> Well,
81:33 | >> I don't know.
81:34 | >> You bring up a good point.
81:35 | >> None of us know the future.
81:37 | >> I don't know. I mean, that's when you
81:38 | get involved and you say, "Hey, we don't
81:39 | want this new upgrade." or
81:41 | >> Well, that's where I was going with
81:42 | that. So, it's like
81:43 | >> So, basically, but
81:45 | with what he's saying, if you're
81:48 | if someone else, whenever, down the
81:50 | line, does go for I want the latest and
81:54 | greatest software and hardware,
81:56 | >> Cuz why wouldn't
81:57 | >> is that
81:58 | something
82:00 | that would then have to be approved by
82:03 | the select board?
82:05 | So, it would be
82:06 | >> It would depend upon how it was
82:07 | purchased. If it was purchased through
82:09 | the police department or whether it was
82:10 | purchased at town meeting through an
82:12 | article, then it would be, you know,
82:14 | depending upon how it was purchased.
82:16 | >> Through the Through the police
82:17 | department, you have the the strong
82:19 | chief statute or whatever that gives you
82:20 | like purchasing authority for these
82:21 | different systems, right?
82:22 | >> Within my budget.
82:23 | >> Yes.
82:23 | >> So, my budget goes through town.
82:25 | >> Yeah, so the budget's is to 9.
82:28 | >> There's also the budget I just can't
82:29 | spend money crazy. I get a certain
82:30 | number. It's like a checkbook. I got to
82:32 | balance the checkbook at the end and we
82:34 | have money left over.
82:37 | You know, one of the things that guys
82:38 | thought about uh
82:41 | and I kind of got a little wary of was
82:44 | you can put these devices on a vehicle.
82:47 | But the problem is is that when they are
82:48 | seriously they're they're hard wired.
82:51 | So, when that vehicle is not used it
82:53 | doesn't help it. So, you got to be using
82:54 | that vehicle
82:56 | specifically driving around. Well, let's
82:58 | be honest. You know, their offices are
83:00 | in doing reports. Like these things like
83:02 | machines like the Terminator. They don't
83:04 | tire. Like they're they're out there
83:06 | 24/7 running and if you're looking for
83:09 | hey, this vehicle was just involved in a
83:11 | you know, whatever
83:13 | you know, a shooting.
83:14 | It's going to indicate
83:16 | to the officers, hey, that just went
83:17 | down Stafford Street. They don't have to
83:19 | know where it went, but they're going to
83:20 | know what time it crossed.
83:22 | Another thing that happens a lot of
83:23 | times in the police department is
83:26 | we work 24/7.
83:29 | Something will happen at 3:00 in the
83:30 | morning and they'll say, "Hey, there was
83:32 | a shooting on on Webster at 3:00 in the
83:35 | morning and they're looking for a blue
83:36 | Honda Civic."
83:39 | Well, 8 hours later it's a different
83:40 | shift. Do you hope they still remember
83:42 | that or if they read the log or they're
83:44 | still as vigilant looking for it? But
83:46 | that what what Flock will do or camera
83:48 | system LPI system is immediately notify
83:51 | that vehicle just went by because it's
83:54 | been looking for this.
83:55 | >> Yeah, now everybody knows.
83:58 | >> Where we live geographically is tough,
84:00 | too, because you can be in Connecticut
84:03 | with under an hour. You can be, you
84:05 | know, much under an hour. Rhode Island
84:06 | in under an hour. So, that's why
84:08 | typically when cars are getting stolen a
84:10 | lot of times they're going across state
84:12 | lines to make it hotter. So, sometimes
84:14 | as soon as we get a hit an indication
84:15 | that a stolen vehicle is coming in they
84:18 | can be, you know, across state lines in
84:20 | a matter of short short order.
84:21 | >> When I when I was on the road working
84:23 | midnights, if somebody called and said,
84:25 | "Hey,
84:26 | I just we they were few and far between.
84:28 | We didn't have a lot of cars stolen, but
84:30 | or more, but people, you know, a nurse
84:33 | coming home at 3:00 in the morning or
84:35 | somebody wakes up to go to bathroom and
84:36 | see somebody stealing a car,
84:38 | we would send if we had
84:40 | available people, we would send somebody
84:42 | to 20 and 80 as close to 84 because we
84:44 | knew the cars were going out of
84:46 | our pattern was they were going out of
84:48 | state.
84:49 | >> So, you still
84:50 | >> do that? But, do you rely on this to do
84:52 | so much on this guy?
84:53 | >> still do that, but
84:55 | it's a
84:56 | we're we have to
84:57 | we're predicting something where is
84:59 | Flock is static and it's there and it's
85:01 | watching like the cars going by it. So,
85:05 | we can focus on other areas
85:07 | like we know if a car gets stolen and
85:10 | it's going to Worcester, the Flock is
85:11 | going to catch it.
85:12 | >> Okay.
85:13 | >> Um if it's going to
85:15 | Connecticut by 31, we know the Flock on
85:18 | Main Street is going to grab it. Okay.
85:20 | Um so, we don't have to send somebody to
85:23 | those areas. We can go to
85:25 | area like the lesser
85:27 | >> Makes it a lot easier.
85:28 | >> probability, right. But, um it does
85:32 | help us. It does like it does increase
85:36 | our ability to solve crime and to be in
85:38 | two places at once.
85:40 | >> That Can you see the Masspike stuff?
85:42 | >> No.
85:42 | >> That's a separate separate issue.
85:45 | >> That's nothing. I don't even know if the
85:46 | state police can see their own see the
85:47 | pike.
85:48 | >> Can we see stuff? Like do you have
85:50 | access to stuff? Can we see something on
85:51 | a camera? I thought we're going to get
85:52 | to look at it.
85:53 | >> Yeah, I can set one up. Sure.
85:56 | >> Anthony, any idea
85:58 | >> I don't know.
85:59 | >> how many stolen cars probably a month
86:01 | come through town?
86:02 | >> I can find it in 2 seconds.
86:06 | It's going to take me 1 second just to
86:08 | log in and then
86:12 | >> Have either of you guys spoken to any of
86:13 | the local folks like you Everybody knows
86:16 | we still got them, they don't if you
86:17 | talk to them, but like
86:18 | >> I have
86:19 | >> Why'd you take them down? What was wrong
86:20 | with your What was wrong with you
86:21 | obviously some places think they're
86:23 | >> So, contract problems and what we got
86:25 | the same problems.
86:26 | >> So, this isn't hurting anybody.
86:27 | >> Be careful in what you see because a lot
86:29 | of places have
86:30 | suspended.
86:32 | >> Yes.
86:32 | >> They're not stopping using it. They're
86:34 | suspending it for right now until
86:36 | the like some issue.
86:38 | >> to calm down.
86:39 | >> Until they're voting. A lot of the towns
86:40 | are voting, right? Well, Southbridge is
86:43 | one that is the same.
86:45 | Yeah.
86:46 | They were
86:46 | >> I just need you to bear with me one
86:47 | second.
86:48 | >> Or you can just put it on pause and you
86:49 | can hit unpause real quick
86:52 | later if you
86:53 | unless you take them down or
86:56 | >> Do you have a comprehensive list of the
86:58 | towns that
86:59 | >> There's 30 plus a couple of weeks ago.
87:01 | >> Yeah, yeah, it's changed I'm sure since,
87:03 | but
87:04 | >> And there's certainly more than 120,000
87:05 | cameras out there or security devices
87:07 | out there.
87:10 | Yep.
87:12 | I might show them to you.
87:13 | >> Okay.
87:17 | >> The camera on 169 does not look 12 ft
87:19 | high, by the way.
87:20 | >> It is. I'm I'm 5 ft 7 and uh
87:24 | it does not look it. I'm telling you
87:26 | that 100%.
87:27 | >> tall.
87:27 | >> And
87:28 | >> The one in front of the library does,
87:30 | too, but the pole is all the way up.
87:31 | >> That one looks tall to me, but when I
87:32 | drove by the 169 one I was
87:34 | >> It's just It's just cuz there's nothing
87:35 | around it for perspective.
87:39 | >> I haven't gotten to walk up by like the
87:41 | Stafford Street one yet. I live right
87:43 | near it so I can pass it like leaving my
87:45 | house every day basically, but is that
87:47 | pole cemented right into the ground or
87:50 | are they installed in like a shearing?
87:53 | >> It's supposed to be a breakaway.
87:54 | >> Yeah, like a shearing plate.
87:56 | >> You know what? I
87:57 | I haven't looked.
87:58 | Uh
88:00 | Give me 2 seconds. Just a second.
88:03 | >> You know I'm sorry.
88:04 | >> How many things can they do at once
88:05 | towns?
88:06 | >> Yes.
88:06 | >> You know what I did read, you can Google
88:08 | it, is um Cambridge. So, you remember
88:10 | Cambridge, they voted to take them down,
88:12 | they had the shooting, and guess what
88:13 | they wanted to
88:15 | Oh, maybe we over overly hastily got rid
88:17 | of these cuz that might have helped.
88:19 | >> Yeah, but those had a lot of false
88:21 | positives with the shootings there.
88:23 | >> I'm sorry?
88:23 | >> There were a lot of false positives with
88:25 | that with that shooting with that noise
88:27 | camera thing, too. Yeah. It was making
88:29 | It was picking up on stuff that wasn't
88:30 | there.
88:31 | >> Oh, you're referring to ShotSpotter.
88:32 | That's uh
88:33 | >> ShotSpotter. Yeah, that's a different
88:34 | system. That's something that So many
88:35 | systems just like
88:36 | >> Yeah. No, you're right. You're right.
88:37 | That That hears noises that are similar
88:39 | to bullets and says, "Hey, uh
88:41 | there was a shooting in this area." And
88:43 | that's So, that's this That's what's
88:45 | come across is they thought that they
88:47 | might have overreacted getting rid of
88:48 | both to Brian's point, ShotSpotter and
88:51 | Flock, because when there was that
88:52 | shooting
88:53 | >> Have you had any conversations with like
88:55 | other police chiefs in the area about
88:57 | it? Have you talked to any of them?
88:59 | >> No. I mean,
89:01 | >> I mean, have they
89:04 | the ones that have disabled or
89:05 | discontinued their contracts? And I
89:07 | mean, have they
89:09 | all like
89:10 | disclosed everything why they did it?
89:12 | >> The most of it
89:13 | >> pressure or was it
89:14 | >> Yeah, I think I think most of them
89:16 | thought uh it just the you know, the
89:17 | juice wasn't worth the squeeze. So, um
89:21 | right now, it's it's a terrible time.
89:23 | Right now, I wouldn't want to be a Flock
89:24 | stock owner. You know, I mean, they've
89:26 | they've gone through the wringer.
89:28 | Um you know, and certainly
89:30 | I don't want to see them ever be used or
89:33 | misused.
89:34 | >> I I wasn't paying attention and I what?
89:36 | >> I was like, "Why is he talking about my
89:37 | family?"
89:38 | >> No.
89:38 | >> Almost nervous.
89:39 | >> I'm almost
89:40 | >> read chiefs text me and saying, "Hey,
89:42 | you at least have the courage to support
89:45 | the what it you believe positively
89:47 | does."
89:47 | >> Look, I mean, you know that I mean, us
89:50 | our concerns is not with you guys. Like,
89:52 | we don't we don't think our
89:55 | >> I brought cookies just so that we didn't
89:57 | have
90:00 | I don't want to
90:00 | >> I think that a lot of this comes from
90:03 | the
90:04 | >> There's a lot of mistrust, don't you
90:06 | >> You know, and you always wonder what's
90:07 | next or
90:08 | >> Yeah, and and that's where this comes
90:09 | from. This is by no means like any every
90:12 | single person I talked to about this
90:15 | there's by no means anyone is
90:17 | saying anything negative about our
90:19 | intentions in this town. It's the
90:22 | it's the mistrust of AI
90:25 | big corporations, data mining, that kind
90:28 | of thing.
90:29 | >> I feel like my friends with the police
90:30 | stickers might not work anymore though.
90:33 | >> Can I have one?
90:33 | >> Yes, they're here.
90:35 | >> Thank you. Hopefully so.
90:36 | >> So, this is what a stolen plate alert
90:38 | looks like.
90:40 | All right.
90:41 | Um
90:43 | picture of the car
90:44 | north side of Stafford Street
90:46 | uh
90:48 | and the license plate um date and time.
90:51 | Let me go into it and see if I can find
90:54 | >> Can you pull up something live? Can we
90:55 | see in front of the library right now?
90:58 | >> We don't have live, but we have
91:00 | >> When how often does it
91:01 | >> It's If something happened
91:02 | >> I'd have to know a plate going by. I
91:04 | can't just live stream uh I don't have
91:06 | the ability to live stream. You want to
91:08 | run my cruiser for going
91:10 | >> Check out I got the copy of the keys
91:11 | already, so
91:13 | >> What's that?
91:14 | I didn't hear you. What'd you say,
91:15 | Brian?
91:15 | >> I said I got a copy of the keys already.
91:17 | I'll just go do
91:18 | >> So, you're the same as everybody that's
91:19 | stealing cars. They uh
91:22 | is a technology out there you can buy it
91:24 | for
91:26 | uh
91:28 | I think it's a less than $3,000 you can
91:30 | clone the car. Like we've had a ton of
91:32 | problems with dealerships. They'll go
91:34 | in, they'll clone a key
91:36 | and the in the morning the dealerships
91:38 | come in and the
91:41 | uh the Dodge Charger is gone. The the
91:45 | pickup truck is gone because they cloned
91:46 | the key in the parking lot while they
91:49 | were there and then left. So.
91:51 | >> I can see how they'd have security
91:53 | cameras. They
92:00 | >> So, just unrelated. So, I am going away
92:03 | to the airport
92:05 | next Friday.
92:08 | So, I'm like, "I better pay for parking
92:09 | before cuz that's a nightmare getting
92:10 | down there and like, how are you going
92:12 | to get screwed?" And you know, so I'm
92:13 | going online and I try to like locate,
92:16 | you know, the my deal for parking. And
92:18 | you literally
92:20 | you put in your plate
92:22 | and it will, when I show up, it's going
92:24 | to the gate's going to open and then you
92:25 | put in the times and, you know, the I'm
92:27 | going to be there for 4 days and it's
92:29 | going to, you know, but literally you
92:30 | put in your license plate and state and
92:33 | license plate number and that's how it
92:36 | and then it gives you a you can also
92:37 | print off a um
92:40 | you know, the code, but but anyways,
92:42 | it's just so they said, "Hey, you just
92:43 | roll in and you just got the gate's
92:44 | going to open." And you've said that's
92:46 | common now at a lot of storage
92:48 | facilities.
92:48 | >> storage facilities
92:50 | are using it. So, this flock has two
92:52 | entities, private and public, okay? A
92:55 | lot of the private stuff that I've seen,
92:58 | um you're on a storage facility,
93:01 | um the gate opens
93:02 | when it read like
93:04 | I'm paying for 3 months of storage,
93:06 | these are the two cars that are going in
93:08 | and it the gate opens up when it sees
93:10 | the plate. Um
93:13 | and
93:14 | the other thing is like, we don't have
93:16 | them around here, but gated gated like
93:18 | condo
93:19 | uh complexes um open up same thing. You
93:23 | have two cars that they're allowed to be
93:25 | there, guest parking is outside and
93:28 | that's what opens up. So,
93:30 | um
93:31 | >> And I think there's a value to something
93:32 | like Amazon when they've got trucks
93:34 | coming in and out, they can probably
93:35 | keep track of what time somebody left
93:38 | when they loaded them up and what time
93:39 | they pulled out of the, you know,
93:40 | there's some stuff that you could see
93:42 | the private sector benefiting from.
93:44 | >> Yeah, I don't really have a problem with
93:45 | like license plate reading and cameras.
93:47 | It's more the system behind it that
93:49 | aggregates all your information and I
93:51 | think more so in like big cities that
93:53 | are using all of the features of the
93:55 | stuff and like that, that starts to feel
93:57 | like that.
93:58 | >> I can't envision a I don't want to
94:00 | envision a town where there's a
94:01 | camera on every corner. I don't. Like I
94:04 | just think
94:04 | >> It makes sense on businesses and things
94:06 | like somebody breaks into your shop or
94:07 | whatever like you have video and like
94:09 | stuff like that. I don't
94:11 | Yeah.
94:12 | You're in public like
94:14 | >> Yeah, I
94:15 | I mean if this technology goes airborne
94:17 | and we only control
94:18 | so much.
94:23 | >> I I work in AI development and it's
94:26 | terrifying. So that's why it scares me.
94:28 | >> That lady that was on the video, wasn't
94:30 | she an AI person?
94:31 | >> No, she was real.
94:32 | >> If somebody walks by that camera, do
94:34 | they get
94:34 | >> I've watched it twice now. You know, is
94:36 | that real or is
94:36 | >> It has to be real.
94:37 | >> I I had to look at it. She was not AI.
94:40 | >> That was uh
94:42 | poor technology for us. So I had a lot
94:45 | of trouble embedding a
94:47 | uh a video.
94:48 | >> So if I go over in front of the library
94:50 | right now and wave at it or put my hand
94:53 | up in front of it, is it going to see
94:54 | anything? You're going to be able to see
94:56 | it?
94:56 | >> No, uh
94:57 | all I'm going to see is the
94:59 | the a vehicle, but uh give me 1 second.
95:01 | I'm just running
95:02 | >> There must be somebody who's driven by
95:03 | in the past 5 minutes here.
95:06 | >> I'm going to drive by while you wave in
95:07 | front of it. I'm going to run the plates
95:08 | and see if it picks it up.
95:10 | So I keep I I used my plate, but like
95:14 | >> You'll be seeing a picture every 60
95:16 | seconds.
95:17 | >> I mean I drove I didn't drive by it
95:18 | today.
95:20 | >> So like
95:21 | >> Oh, thank you.
95:22 | >> You know what I mean? Like that's a
95:24 | Stafford Street.
95:25 | Uh
95:26 | behind Ted's. Like that's my cruiser.
95:29 | That's
95:30 | if we're searching for my cruiser.
95:32 | >> Is that at the fire department or at or
95:33 | the Ted's property?
95:35 | >> No, it's behind it's on Stafford Street
95:36 | behind.
95:38 | So
95:38 | >> I didn't look for it, but I heard it was
95:39 | Ted's property.
95:41 | >> I'll give you another example. The
95:43 | closest thing we could do
95:45 | to searching
95:47 | uh
95:50 | Give me a second here. The closest thing
95:52 | we could do to searching is
95:54 | give me a vehicle
95:57 | >> We can search by anybody's plate that
95:59 | drove by here today, right? Anybody
96:01 | drive by the camera an hour ago?
96:03 | >> I did.
96:04 | >> No, let's I don't want to because we're
96:06 | on cable access, I don't want to put
96:07 | somebody infringe somebody's
96:10 | rights.
96:11 | >> Like a red Camry or something.
96:13 | >> All right, body style. Do you want to
96:14 | sit in?
96:15 | >> Sure.
96:20 | >> Color?
96:21 | >> Red.
96:22 | >> I don't know.
96:23 | >> White, common color.
96:29 | >> You know what? I'm going to leave it all
96:30 | just to show you.
96:40 | >> Thank you all very much. I appreciate
96:42 | it.
96:43 | >> So, this is what we have.
96:45 | That's red sedan, okay?
96:48 | Um
96:49 | This is as close as we can get
96:55 | to real time. Like I can't
96:57 | >> That's That's fine.
96:58 | >> Um I can't get any
97:00 | >> That's 1,400 results in like 30 days or
97:03 | whatever it's going back 30
97:04 | >> Um I I only went back a week.
97:08 | I did 7 days. So, just to give you an
97:10 | idea, that's
97:12 | it like I can't randomly search like
97:15 | I don't have the capability to do some
97:16 | of the stuff.
97:17 | >> So, is each of those pictures just
97:18 | attached to one car?
97:20 | >> Like there's presumably like six little
97:21 | pictures there or are those all
97:23 | sequential six pictures I'm looking at?
97:25 | >> No, those are different cars.
97:26 | >> Yeah, but there's there's multiple
97:28 | pictures of each of those cars, right?
97:30 | >> No.
97:31 | >> So, so just one picture of that car.
97:33 | >> Just one picture of the car cuz it got
97:35 | it got what it's looking for, the
97:36 | license plate.
97:36 | >> That's my daughter.
97:38 | >> Is it really?
97:39 | >> Yes.
97:41 | >> She told me she was home.
97:43 | We're going to split a lottery ticket
97:44 | tonight.
97:49 | >> Um my god, no.
97:53 | >> I hope the plate comes up right now.
97:55 | >> I can't remember her license, but it
97:56 | could be
97:57 | >> If we did that, wouldn't it
97:58 | >> That would be weird.
97:59 | >> Then there's something else.
98:01 | >> Yeah, her car's not Her car's a room, so
98:04 | >> So, like, presumably you drive around
98:06 | all four of the cameras, and then I
98:07 | can't
98:08 | Oh, I passed all four of those cameras
98:10 | in some amount of time, and I've
98:11 | probably traveled this path to pass all
98:13 | four of our cameras in town.
98:15 | >> I I'll tell you this, be a lot of effort
98:16 | on you to go do that in this system, it
98:19 | looks like.
98:19 | >> Yeah, you couldn't just
98:21 | like it would
98:23 | it wouldn't pick it up. Like, I tried it
98:25 | with somebody that came in.
98:26 | If people have been asking for
98:28 | demonstrations, they couldn't make the
98:29 | meeting, we did some stuff with them.
98:31 | Um I actually had an officer drive
98:32 | around in his car
98:34 | so that we could show, "Hey, this is
98:36 | what it it does." cuz I knew his plate.
98:38 | Um
98:40 | the only
98:41 | Sorry, this week it's the crappy
98:42 | picture.
98:44 | If I did pick your daughter out, then
98:45 | we're going to buy a lottery ticket, cuz
98:47 | I literally picked just a random
98:49 | >> I don't
98:50 | She What is that the last 7 days?
98:53 | >> Yes.
98:53 | >> She's home from school.
98:55 | Go get her.
98:57 | I don't remember. I can't I'm a woman, I
98:59 | don't remember license plate numbers.
99:02 | >> All right, so
99:02 | >> You catch different ones every
99:03 | >> You guys ask some questions before, real
99:05 | quick.
99:07 | It's it's getting there.
99:09 | All right, so total volume of cars
99:12 | year-to-date volume, 5 million cars have
99:14 | been read
99:15 | in town.
99:16 | >> And those are non-unique, like it could
99:17 | be like my car multiple times.
99:19 | >> Yep, 100%. Okay.
99:21 | >> And you know, to your point, 40 times a
99:23 | day, you know, if you're going back and
99:25 | forth.
99:25 | >> Yeah, yeah.
99:25 | >> It just
99:26 | >> Like, just I'm going to scroll down. I'm
99:28 | limited a little bit by the
99:32 | by the screen, but like
99:35 | today, 23,000 cars
99:38 | in town that hit a flock, okay? There's
99:41 | a lot of traffic here.
99:43 | Um
99:43 | >> Oh yeah, I live right off Stafford. It's
99:45 | constant.
99:45 | >> Yeah.
99:47 | Forget it. You don't try to cross in
99:48 | Main Street.
99:49 | Main Street at 5:00 not fun.
99:56 | Or between like 2:30 and 4:00 across the
99:58 | school.
99:59 | >> Yeah, the bus is on.
100:03 | >> I've been here a long time. When I first
100:07 | started going to work
100:10 | I would just turn
100:12 | pull onto Main Street.
100:14 | Now, sometimes I'm waiting up to 5
100:17 | minutes
100:17 | >> Oh yeah.
100:18 | >> depending on when.
100:19 | But I don't work anymore, so that's
100:21 | okay.
100:22 | >> I should have tried to pass 169.
100:26 | When there's problems on the pike.
100:27 | >> Yeah, it's bad everywhere.
100:31 | >> Sorry, I'm trying to get a
100:32 | >> You never go Route 20 east
100:35 | >> No.
100:35 | >> on Sunday, well, any day around 3:00.
100:38 | >> Yeah.
100:40 | >> And you don't do it
100:42 | at the 4:00 or whatever on Sundays of
100:45 | the weekend people.
100:49 | >> And you go down to like the Walmart down
100:50 | there or whatever it's strawberry and
100:51 | you try to come back and you're stuck
100:53 | for 45 minutes and I try
100:55 | >> Well, no, that's when you just
100:57 | go and drive drive to
101:01 | >> Okay, so year-to-date
101:03 | stolen vehicles
101:05 | >> from 169
101:07 | >> 1,082
101:09 | um
101:11 | Sorry, stolen plates, stolen vehicles uh
101:13 | just in town 386. Sex offenders
101:16 | um 129
101:19 | gangs
101:20 | uh suspected terrorists have come up on
101:22 | an NCIC hit um 23
101:25 | >> So, how do you get these sex offenders
101:26 | and the terrorists? What, they just by
101:28 | their license plate?
101:29 | >> They have to register uh sex offender
101:31 | when they register in Massachusetts has
101:33 | to register with their license plate and
101:36 | any vehicle that they drive. So, if you
101:38 | go look at like our published
101:41 | uh statewide. Uh so, if you go on the
101:43 | Sex Offender Registry Board,
101:44 | um they have to provide their vehicle,
101:46 | they have to provide a a whole bunch of
101:48 | information as a condition of
101:49 | registration. Since they
101:50 | >> And certain levels are they
101:51 | automatically tracked as they come
101:52 | through town?
101:53 | >> They're not. No, it's mostly
101:54 | >> It says how many are up there, right?
101:55 | With
101:56 | >> No, so at certain levels. So, certain
101:59 | levels don't have to provide them,
102:00 | certain levels do, and then people that
102:02 | are in violation
102:04 | um
102:05 | who are traveling through. And it might
102:06 | be another state.
102:07 | >> They get like warrants and stuff, too?
102:09 | >> If you don't register If they don't
102:10 | register
102:11 | >> not you're not for the sex offenders
102:12 | specifically, for
102:14 | just just people with warrants or
102:15 | >> Yes, so like uh
102:17 | 14 warrants, we
102:19 | uh typically don't use it for warrants
102:22 | um
102:22 | because
102:24 | >> That's just Dan's point though.
102:27 | With a warrant, it's only if the I could
102:29 | have a warrant and drive somebody else's
102:31 | car and it's not going to indicate. It's
102:33 | only if I have a warrant and that car is
102:35 | somehow through law enforcement or
102:36 | something that's been attributed to me
102:38 | that they can put it in that file so
102:40 | that it's a hit to say, "Hey, there's a
102:42 | likelihood." But you
102:44 | you could have switched cars. You When
102:45 | you stop that car, that doesn't mean
102:47 | that person has a warrant. That just
102:49 | means there's a high likelihood
102:50 | >> You know that somebody with a warrant
102:52 | drove by you.
102:52 | >> The chief makes a good distinction there
102:54 | because
102:55 | um these are warrants in NCIC. So, these
102:58 | are warrants It doesn't just run WMS
103:00 | Massachusetts warrants. It runs vehicle
103:04 | ones that are linked to a person with a
103:05 | vehicle.
103:06 | So, it's and NCIC warrants are a higher
103:10 | degree There's more serious crimes.
103:12 | >> So, Flock integrates those
103:14 | >> Because they're nationwide. Those are
103:15 | run nationwide.
103:16 | >> So, you don't have to input like
103:17 | >> Those are
103:18 | >> the sex offenders of the town into here
103:20 | >> Those are at the
103:21 | uh federal level, NCIC.
103:23 | >> Do you have terrorists that come up also
103:25 | or just
103:25 | >> Um
103:26 | >> Is that what one of the lines was?
103:28 | >> Yeah.
103:28 | >> Yeah, but they yeah, uh gang activity at
103:31 | that's entered through NCIC. These are
103:33 | all NCIC. These are not things we're
103:35 | entering.
103:36 | Um
103:37 | but
103:39 | uh and we've had a missing person
103:41 | um
103:45 | just looking at my date uh
103:50 | hit there.
103:51 | Um
103:52 | >> So, this is so much better and worse
103:54 | than I thought.
103:55 | Like, there's a lot of positive aspects
103:57 | to it. There's a lot of cool things to
103:58 | it.
103:59 | >> good and there's bad.
104:00 | >> Almost everything Anything you do, I can
104:02 | tell you there's a pro and a con.
104:04 | Um and a lot of times when people will
104:05 | say, you know, like ICE is a hot-button
104:07 | topic. Again, this is a license plate
104:10 | reader. You'd have to have, you know,
104:12 | when we get like a a hit that there's a
104:14 | crime occurring, like we don't say, "Oh,
104:16 | there was a shooting down the in
104:18 | Southbridge or Worcester.
104:20 | What's their immigration status?" Cuz
104:22 | we're not going to stop them if they're
104:23 | this or that. Like, there's a crime.
104:24 | We're going to stop them. But, we don't
104:26 | take any federal action on somebody's
104:29 | immigration status. But, if there was a
104:31 | a shooting and that car had to be
104:32 | stopped whoever the suspects are in the
104:35 | car would get out, possibly arrested and
104:38 | then
104:39 | and I just want to say the immigration
104:41 | But, we're not hunting for immigration
104:43 | for somebody that's overstepped their
104:45 | visa or something like that.
104:46 | >> This is what the federal government's
104:47 | doing that. And you're You're providing
104:50 | >> I don't care what the federal government
104:51 | is doing, but I would tell you this
104:52 | probably.
104:53 | This is a license plate reader, not a
104:55 | person finder. You know, I mean, you
104:57 | might find a person that's driving a
104:58 | registered car if they had a warrant
105:00 | that was entered. But, it's not going to
105:02 | tell me if I don't know what car you
105:03 | drove here tonight.
105:05 | >> Yeah, but most most usually people are
105:07 | operating their own vehicles.
105:08 | >> Usually No, you're You're not wrong. I'm
105:10 | just saying that I mean, the chances are
105:11 | you could have driven something else
105:13 | here and you know, it is what it is.
105:14 | It's just going to tell us what vehicle
105:16 | went by at a certain time and place.
105:18 | >> I'm glad it's you guys at the helm of it
105:20 | because like this system is capable of
105:22 | so much more and other departments that
105:24 | might not be as
105:25 | >> You guys seem reasonable and respectful
105:27 | it.
105:27 | >> We appreciate that.
105:28 | >> Like I seem reasonable. I don't know
105:30 | what this
105:31 | >> That would have That would have looked
105:32 | too funny.
105:33 | >> The fact that the system can do it is
105:34 | the terrifying part.
105:35 | >> So, one other thing I talked about, and
105:37 | I haven't I've intentionally not cleared
105:39 | these out just so I I knew people are
105:41 | asking about them, okay? This is an AI
105:44 | audit function, okay? Remember I showed
105:47 | you before the Excel spreadsheets? So,
105:49 | this condenses it. It see
105:51 | So, you see searches beyond a 30-day
105:53 | timeframe.
105:55 | I intentionally
105:56 | tried to push a search beyond 30 days
105:59 | to show people I can't do it, okay?
106:02 | Um
106:04 | and it flags, "Hey, why is he He knows
106:07 | it's 30 days.
106:09 | Why is he searching beyond the 30 days?"
106:11 | So, it says
106:13 | it sends us a a
106:14 | a notification.
106:16 | Um
106:18 | and that's my cruiser that I
106:19 | intentionally tried to push beyond 30
106:21 | days, okay?
106:23 | Um
106:25 | Can I ask you not to hit the screen for
106:27 | a minute with the camera? Can you
106:29 | >> I'm below it right now, so it's not on
106:31 | camera.
106:31 | >> Okay.
106:32 | Um
106:36 | the
106:40 | Give me a second here.
106:45 | All right, so
106:47 | remember the ones I was telling you
106:48 | about with repeated narrow timeframes
106:50 | where somebody's hitting refresh?
106:52 | It's It's catching those, okay? That's
106:55 | somebody doing good police work. They're
106:56 | sitting on Route 20 or they're sitting
106:58 | somewhere where there isn't a camera,
107:00 | and they're using the camera to watch
107:03 | the other street so that they can find
107:04 | the person um
107:06 | and
107:07 | Flock We We talked to Flock about this.
107:10 | They said, "Hey, there's a
107:11 | custom feature
107:13 | in the program that we can set up so
107:15 | that it can do that without triggering
107:17 | this audit."
107:18 | >> I was going to say
107:18 | >> It will record it all, but it doesn't
107:20 | trigger the audit, okay?
107:23 | And I mean with anything there's things
107:25 | we learn
107:26 | like as we we use it, okay?
107:28 | >> one who does the audit? Is it
107:30 | >> I do it and I give it to Deputy Mason.
107:33 | So.
107:34 | >> An audit's not bad things. Keeps
107:36 | everybody honest. Like you know, if
107:39 | people know, hey, people look at this
107:41 | and and by the way, the system looks at
107:43 | this, too. So, there's two sets of eyes,
107:46 | the system and a you know, a supervisor.
107:50 | >> Um
107:52 | but that's that's like where it shows
107:54 | you that it's
107:56 | um
107:57 | the AI function. And I I haven't um
108:02 | I haven't cleared those out
108:03 | intentionally so that I could show them.
108:06 | Um
108:07 | >> Can you do um bumper sticker?
108:09 | >> Yeah, we can do that.
108:10 | >> I guess you said like there's a there's
108:12 | a false narrative out there that
108:15 | that you're looking for whatever. Say
108:17 | you're you're anti-Trump or you're
108:19 | anti-Harris or whatever and like somehow
108:22 | the police are going after these people.
108:26 | It doesn't bump a sticker though.
108:28 | >> It doesn't specifically look for a Trump
108:30 | or a Harris. It looks for a sticker. So,
108:33 | if he does a search
108:36 | of looking for maybe the person the sus-
108:38 | excuse me, the the victim says, "I don't
108:40 | know what kind of car it was, but it had
108:41 | a bunch of bumper stickers on the back."
108:44 | Then the officer could just look for
108:46 | bumper stickers.
108:48 | Now, it may not be a bumper sticker.
108:50 | Okay, it could be a something else, a
108:51 | decal or
108:53 | but anyway, something that the the
108:55 | software system recognizes that didn't
108:57 | come from the factory with that car.
108:59 | Like a roof rack or something that was
109:01 | put on later. So, that's why that
109:03 | ability to I'm sorry.
109:04 | >> You can't search the content of the
109:05 | sticker.
109:06 | >> Right.
109:06 | >> So,
109:10 | >> your daughter's car again.
109:11 | >> Yeah.
109:14 | >> I can't believe mean if it is to be
109:16 | honest with you. The the randomness
109:18 | All right, so like
109:19 | >> That's a lot.
109:20 | >> Here's a random sticker, okay?
109:23 | >> bit for sure.
109:24 | >> But you can't like we can't search by I
109:28 | picked stickers.
109:29 | Do you see where it just
109:31 | randomly grabs a sticker.
109:34 | Um
109:35 | >> But that might might be the victim's
109:36 | only thing was, "Hey, I don't know what
109:37 | the car looked like so fast, but it had
109:39 | a bunch of it had a front headlight out
109:41 | or it had a
109:43 | Yes, stickers on the back.
109:45 | >> Okay, you know, I don't know what
109:46 | stickers. Okay, it was red. We'll look
109:48 | for a red car.
109:49 | >> search criteria, not necessarily license
109:51 | plate. We can search color, make, and
109:53 | sticker. Color, make, and year.
109:56 | >> You need a couple identifiers. So like
109:58 | if you search just red,
110:00 | you're going to get it's going to
110:01 | overwhelm you, okay? If you get red
110:03 | sedan or you get red hatchback or you
110:06 | get red with a sticker, you need
110:07 | something to eliminate out
110:09 | other stuff.
110:10 | >> Yeah, like this search it went from like
110:11 | 1,400 down to six by adding stickers.
110:14 | >> Uh but I I
110:15 | walked out.
110:17 | Um
110:19 | I I kept it as a red car, sedan, you
110:21 | know, I narrowed it down. Um we had a
110:23 | hit-and-run where somebody was seriously
110:26 | seriously injured and the only thing we
110:27 | had, white car.
110:30 | We had a discrepancy at We knew it was a
110:31 | white sedan and it had a big sticker in
110:33 | the back window.
110:35 | And it took us weeks to figure it out,
110:37 | okay? And the person was like serious
110:39 | injury serious injury to the point where
110:41 | we were concerned they were going to die
110:42 | at first. Um but that's all we had to go
110:45 | with and we didn't have this system
110:46 | then. Um
110:49 | but I mean it does not give you
110:52 | um a whole lot to search other than
110:55 | that it's got a sticker.
110:56 | Um
110:58 | like
111:00 | >> You can't see You can't see the person
111:01 | driving.
111:02 | >> see the person in there.
111:03 | >> No.
111:03 | >> That's the best resolution you got.
111:05 | >> You know how like when you use your
111:06 | iPhone, it's got that yellow box?
111:09 | So the yellow box is looking for that
111:12 | plate. That's what it's it's aimed at
111:13 | that for an angle. That's why they use
111:15 | 12 to 14 ft. Um and that's
111:18 | >> possible right conditions and light and
111:20 | stuff you can get a picture.
111:21 | >> It's just
111:22 | luck.
111:23 | >> Yeah.
111:23 | >> Like it's literally cuz it's taking so
111:25 | many pictures so quickly, you'd have to
111:27 | you know, you could turn and not be
111:28 | facing I suppose, but
111:30 | >> Um
111:31 | >> generally we can't really tell who is
111:33 | driving. You know, people go up there
111:35 | like that's the car and then a lot of
111:36 | times that's the follow-up
111:37 | investigation.
111:38 | >> Yeah.
111:39 | >> If someone is in that passenger seat of
111:40 | the opposite side of the road, like you
111:42 | could probably see their face in that
111:43 | picture.
111:44 | >> They're not in the But you couldn't make
111:45 | them out.
111:46 | >> But you might get enough
111:48 | >> But some software enhancements can kind
111:49 | of pick can can
111:50 | >> I mean that we don't have them sitting
111:52 | there.
111:52 | >> So you can't do it, but some but
111:54 | potentially somebody else has.
111:55 | >> I'm sure the CIA could do a lot.
111:57 | >> They can do it from space now, right?
111:59 | >> I mean they're very able to do it and it
112:00 | is is pretty much
112:01 | >> build rockets for a living, so.
112:02 | >> Oh,
112:03 | wrong three letters to choose cuz
112:05 | they're not supposed to operate in the
112:07 | US.
112:09 | >> I'm going to try to look for a head-on
112:10 | picture and I'm not getting one to give
112:13 | you an idea of like the quality of an
112:15 | image.
112:16 | >> Well, it seems pretty good.
112:17 | >> Now, sometimes like what's helpful is if
112:20 | like on 169, if you're looking for a car
112:23 | at 3:00 in the morning, it's such a
112:25 | small pool of cars because you know, but
112:27 | now if you're looking at you know, 8:00
112:30 | in the morning when everybody's
112:30 | commuting to work, you might have
112:32 | hundreds of cars that go by with bumper
112:34 | stickers and things. I'm that's
112:35 | exaggerating a little bit, but versus
112:37 | you know, if 3:00 in the morning they
112:39 | went by with a bumper sticker, well,
112:41 | you'd probably be able to find it pretty
112:42 | quickly.
112:43 | >> That's the same camera I showed you a
112:44 | minute ago with the car coming at us
112:47 | and you can't see the driver.
112:48 | >> Yeah, I know.
112:49 | >> It is focused the primary image is the
112:51 | license plate.
112:53 | But I do think it's possible that you
112:55 | could get
112:55 | >> You could see a person. I mean,
112:56 | depending on where you are when it
112:58 | snapped the picture, you might, but you
113:00 | might not.
113:01 | But with you know, with slightly
113:02 | advanced cameras, you know, the next the
113:04 | next range whatever probably probably,
113:07 | but not currently.
113:09 | >> Well, I appreciate
113:10 | >> Yeah.
113:10 | >> all the information and all that.
113:13 | >> If you come up with any If you have any
113:14 | other questions, let