Flock Safety LPR Tested

By Derek Ward, Published Jan 06, 2021, 10:56am EST

LPR startup Flock has raised over $80 million, claiming their license plate readers "stop crime", but how well do they work in the real world?

IPVM Image

We tested Flock Safety's Raven LPR camera for over two weeks, examining the following:

  • How accurate are license plate reads overall?
  • How well does it handle increasingly common stacked characters and specialty plates?
  • How accurate is US state identification?
  • How accurate are vehicle make and type identification?
  • How does it handle multiple plates in the scene?
  • How does it handle partial reads?
  • How does the Flock Safety Cloud interface perform?
  • How well does the solar panel work to power the system?
  • 10-minute video review of the Flock UI and system

Executive *******

** *** *******, ***** Safety ********* ********** *** rear ******* ****** ******, day ** *****, **** cars ********* ******* ~** mph (~** ***) ** to ~** *** (~*** kph) *** ********* **** stacked ********** *** ********* plates ***** **** **** issues *** ******* ** past *****. ************, ***** did *** **** ***-***** text, *******, ** ****** as ******, ***** ******* false ***** ** ***** LPR ***** *** *** cloud ** *** **** to ******** *** ***** searches, **** *********, *** list ********.

*******, *** ****** ********* misclassified ******* ***** ***** (e.g., ************ ********** ** Maryland), ******* **** (*.*., mistaking **** *** *****), and **** (*.*., ********* Toyota *** ******). ************, the ****** ** ********* pole ***** ****, **** no ****** ** **** mount, ***** *** ** inconvenient ** **** ************ with ***-******** **********.

***** ***** ***** *** not *********. ** *** testing, ***** ****** **** rarely **** ********* ****** the *** *** **** at *****.

**** ** ****** *** supported.

Video ********

Flock ******* ******** *** *******

***** ******** **** ***** staff ** ******** ** every ******* *** ******** install, ***** ** **** setup, *** *************. ***** markets ******** ** ***-*****, who **** ********* **** the ****** ********* ** implement *** *********, ****** some ***** *** ****** to ****-******* *** **** is *********** (****** **** is ******** ** *** cost).

Accurate ***** ****** *****

**** *** ****** ** 2+ ***** ** *******, Flock ********** ********** **** license ****** ******* *** camera, **** ******** *********** of ******** ******, ***** below.

IPVM Image

******** *** ******* ** night, **** *** ****** using ****** ** ************ when ******** **** ** properly ********** ******.

IPVM Image

Two **** ********

***** ********* ***** ******** width ** "*.* *****" of *******, ******* * single ****** *** ***** two *****, ** ******* plates *** ********* *** present ** *** ***** of **** ****. **** show **** ** ***** documentation, ********* *** ******'* coverage ****** *** ***** with *** ******* **** of **** ** ****:

IPVM Image

Non-Plate **** *******

***** ****** ******* ***-***** text, **** ** ********* on *** ***** ** vans/trucks ** **** ** shadows, ********, ** ****** which **** *** **** issues *** **** ******* in **** *****.

IPVM Image

Multiple ****** **** **************

**** **** ****** **** multiple ******* ****** ** the *****, ***** *** able ** ********* **** each ** *** ****** separately.

IPVM Image

Stacked ********** *** ********* ****** **** ********

******* ***** ********** **** read ** ***** ****** from *** ** ******, and ***** **** ** right. ******* *** ******* characters **** ** *** left ** *****, **** were **** ********* *** included **** *** ****** plate *******.

IPVM Image

No ******* ***** ***** / ***** ********

***** **** *** ******* a ***** ** ********* partial ***** **** ****** in *** ********, ********* in ******* ***** ***** viewed ** ***** **** captures. *** *******, ** the **** ***** * bike **** ****** **** of *** *****, ********* in **** **** ********** being ****. ** *** right, ** **** ******* two **********. **** ** these ****** *** ***** to * *********, *** neither ** ******* *** to ******* *****.

IPVM Image

Incorrect ***** ***** ******

***** ********** ******** ********* license ***** ******, **** when ****** **** ******* in ****. ** ***** plate ***** *** ********* on ***** * ** 10 ***** ****** *** testing.

IPVM Image

Multiple **** ******

****** **** *****, ** noticed **** **** ****** would ** **** ******** times, ***** ***** **** to **** ******** ******* to **** ******* ****** a ***** ******. **** this ********, ** *** common ** *** *** vehicle *************** ******, ***** could **** ******** ** vehicle ****** *********.

IPVM Image

Incorrect ******* **********

****** *******, ** ******* frequent ********* ******* ******, such ** **** ****** SUV ***** ********** ** a "***".

IPVM Image

*******, *** **** *** regularly ********* ** *** recognized, **** ** *** Chevrolet *****:

IPVM Image

***** ****** **** **** frequent ** *****, **** vehicle ********** ***** *** recognized ** *** ** low *****, **** *** plate **********:

IPVM Image

Front ***** ***** *** *********

***** **** *** ******* front ******* ***** ***** by ******, **********:

*** ****** ** ************ designed *** ********** ** capture********** ** ********, ***** is ********** ****** ******. They **** ******* ***** plates *** *** *** designed ** ******* ** the **** ****/********.

*** ******* ******** *****'* response, **** ********* **** accuracy ** ***** ******, even **** ******* ******* during *** ***.

IPVM Image

US ****** ****

***** ****** ********* **** works **** ** ******. European ****** **** *** accurately **** ** *** tests.

IPVM Image

Flock ****** ***** ********* ********

*** ******** ** *** Flock ******'* ***** *********** include:

  • ***** ******: **** ****** ** to *** ***** ***, users *** ********* **** the ******* ***** ****** page, ********* *** ********* search ******** **** *** be **** *** ******* and/or *********.
  • ********* *************: ** **** ***** hits, ***** *** ***** on "******* **** ******" and *** ******* ******* such ** *****, **** and ***** *** **** used ** ****** *******, which ***** ** ****** for ********* **** ** a ******** **** ******* of * ******** *****.
  • ****** *************: ***** ****** ******** a ***** ************ ***-** in *** ****** ** the ****** **** **** alerts ***** ** *** updates ** *** ***********, such ** ********* *** other ********.
  • **** ********: **** ******** *** whitelists (***** ***** ****** classifies ** "*********") *** be ******* *** ****** either ********* ** *** import ** * .*** file. ***** *** **** exported ** .*** *****.

*** ****** **** ****** the *********** **** ***** the *** ******* (****** ~5 ******), **** ** email **** ******** *********** such ** ***** *********, the ****** ******, **** number, *** *** ****** why *** ******* ***** is ** * *******.

IPVM Image

Watermark ****** *****

***** *** ***** ***** the "*****" ********* ***** potentially ***** *** ******* plate ***** ****** * search. ** ***** ***** about ******** *** *********, and **** **** **'* possible *** **** ** do ** ****** *** needed *** *************, ****** there *** ** ******* for ***** ** ** so.

IPVM Image

Physical ********

** *** ******** ********, we **** * **** at *** ******** ************, features, *** ************ ****** of *** ***** ****** camera *** ***** *****. As *** *** ***** panel ******, ***** *** not * *******, ** there *** ********** ******** in *** **** *****, and ****** *** *******, there **** ** ****** with *** ****** ******* off.

***** ******** ****-******* ************. Flock ********* *** **** do *** ********* **** mounting:

** ********* **** *** mount ******* ** * wall/side ** ********, **** for ******** ******* (**** of ******/********* ************* **** holes), *** **** ******* it **** **** ** limit *** *****/******* **********.

*****

*** ***** ** * typical ***** ****** ****** for "************" *** ** $2,000 *** ****** *** year, ** *********:

******* ** ****** ********-*******. For*************, ** * ***-**** contract, **’*$*,*** / ****** / year. **** ******** ********** ************, ***********, ******* hosting, ******** *******, *** software *******.

*** "************" ****** **** includes * ****** "********" list, *** ** ***** configurable ********.

***** ** **** * "police-grade" ****** ***** ******** hotlists *** *********** ** government ********, ***** ** priced ** $*,*** *** year.

*********-***** ****** ****** ** $2,500 / ****** / year**** * ***-**** $*** installation ****. **** ******** ***********, ******* *******, cellular *******, ********* ******* alerts, *** ******** *******.

Comments (36)

**** ***** , **** done ****

Agree: 1
Disagree
Informative
Unhelpful
Funny

***** ******* *** **** this ** * ****** Xterra *** * ****.

IPVM Image

Agree: 3
Disagree
Informative: 3
Unhelpful: 1
Funny: 1

* *** **** ***. I ***** ** ***'* be *** **** ** Flock ************** ** **** also *************. :)

Agree: 1
Disagree
Informative
Unhelpful
Funny: 8

***** *** ***** *** UD #*, **** *** changed!

Agree
Disagree
Informative: 1
Unhelpful
Funny

** * **** *****, I ** ******** ** this *****************!

* **** ** ** Xterra's...

Agree
Disagree: 1
Informative
Unhelpful: 1
Funny: 10

****** **** ** *** why **** **** ** an ********* **** * was **** ****** ** be *******. * ***** hope **** *** **** team **** ** ** me **** ****** ** help **** ******** ***** high ********* *** *** as ** ****** *** at **** ** ****** to ** *********. ** you *** ********* ** on * ****** ****** for ********* **** ** something **** **** ** simply ***** *** *** chose ** **** ********* it ***** ** ***********.

Agree
Disagree
Informative: 2
Unhelpful
Funny: 1

#*, ****** *** *** help. * ***'* **** why ******* ***** **** comment ********* *** * would *** ***** ***** it. * **** **** you ** *********** **** to ********** ****.

Agree
Disagree
Informative
Unhelpful
Funny

******, ***** *** ** worry * **** ***** I **** **** *** 1 ***** ** *** last *** ***** * was **** **** ******* as ** *** * could ******* ** ************* skills ** *** ****** or ******** *** ********** of ****** ****** ** the ******. ***'* ******* see * *** ** frowns, ****** **** ***, so *** ******* ** to ***. ***** *** again.

Agree
Disagree
Informative
Unhelpful
Funny

******** * **** ***** upset **** **** *** now * **** ** Peugeot?

Agree
Disagree
Informative
Unhelpful
Funny: 2

**** ***** *** ********* plates **** ** *******, and *** *** *** system ** ** *****? Our ********** ** *** market (**) ** **** the ******** ** *** asked ** ***** *** most **** ********* ******. And ***** **** ****** are ***** *** *** reflective ** **, *** same ******** **** **** metal ****** ** ***** fail ** **** ***** plates, *** **** *****.

Agree
Disagree
Informative: 1
Unhelpful
Funny

***** *** ***** ******* on *** ****** **** will **** **** *** paper ******, **** ** night. **** ** *****.

Agree
Disagree
Informative
Unhelpful
Funny

******** **** ****** **** at **** *** ***** plates.

Agree
Disagree
Informative
Unhelpful
Funny

***** ** #*,

****** *** **** ******* the ******, ***** **** only * ******* ** temp ****/****** ****, ****** during *** *******, **** were ********** ******* *****.

IPVM Image

** *** *********, ********* temp **** **** **** accurate, **** *** ********** reads ****** *** ***/**** tags, ** **** ****:

IPVM Image

Agree
Disagree
Informative: 2
Unhelpful
Funny

***** *** *** *** detail. **** ********** ** what * ***** **** guessed. ** **** *** a *** ** ***** vehicles **** **** (****-*******), damaged/faded ***** ****, *** paper **** ****** ** plastic ****. *** ******* bags (** **** **** dry ** *** ****) are *** *****, ** the ***** **** ******* off ******* ** *** bag, ****** **** *** and ***** ****** *********.

Agree
Disagree
Informative: 2
Unhelpful
Funny

***, ****'* ****! ** terms ** *** ******** bags, *'* ******** **** are *** ***** ********-***** baggies *** *** *** grocery ***** ******* *** variety, ******* ** ****'* the ****, * ***'* know **** *** ****** would *** *****.

****** *** *** ******-** info!

Agree
Disagree
Informative
Unhelpful
Funny

**** ** ** ******* from ******* *****:

IPVM Image*** *** *********** **** put *** ****** ** these ******* ******* **** inevitably *** ******** ** creased **** ****. *** individuals ****** ** ***** own ****** (*** *** find * ******** *** anything ** *** ********) will *** **** ** Ziplocks ** ********* *******. Now ******* ****** ****** sun ** *** *******, or **'* *****. ****** usually ***'* ****** ***'* or ********* ***** ***** valid *****. *** *** list ** ****** ***** plates, ** **** **** (fortunately) ***** **-*** * paper ***** ** ******** crimes.

Agree
Disagree
Informative: 4
Unhelpful
Funny

**** ******** **** **** on *** ****** (******* speed, ****, ***)? **** Flock ***** *** **** to ****** **** ** is **** **** ***** in?

Agree
Disagree
Informative
Unhelpful
Funny

** *****,

****** ******** *** **** coded, ** ***** *** no *** *** ** to *** ** ** it *** *** **** was ***** ****. ***** are ** ****** ******** to ***** ** ***** Safety ***** *********.

Agree
Disagree
Informative: 1
Unhelpful
Funny

*****, **** *******. *** you **** ***** **** of *** ********** ****** this *******. * ***** see ** **** ****** you **** ******** ****** this *** ****

Agree
Disagree
Informative
Unhelpful
Funny: 1

**** ***** **** ** PII ********** ****** **** testing

** *** **, * do *** **** ** any ***** ***** ** capturing ******* ****** ** we *** *****.

** *** *** ********** in **** ** ******** on **** ***** *** some ** *** **** controversial *********, **** **** -**** ******* ***** * Private ************ *******. ** Tracked ******* **** **

Agree
Disagree
Informative
Unhelpful
Funny

**** - **** **** as ******. *******, * did **** * ******** about ******* *** *** any ******* **** ***** smoke-colored ******* ***** ****** that *** ********* **** frequently. ** ***** **** the ***** **** ***** for "***** ******* ******* Make ******** **********" *** a ***** ******* *****. I *** ******* ** had ** *** **** was ******* ** **** they *** **** ******* our ************ ** * DW *** ****** *** their ************ *** ** were ************ ******* *** of ***** ********'* ***** on *** *******. ***** extensive *******, ** ***** out, **** *** * fairly *** ******* **** a *****-******* ***** *** the ****** **** ******'* see ******* ** ** night. **** * ****** it **, * **** found **** ** **, that **** ** ***** cover ** ******* *** I'm ******** ** *** them ** ********** ***. If *** *** *** chance, *'* ** ********** to **** *** *** Flock ******* ******* **** that **** ** ***** and *** **** **** to ******* *** ***** correctly.

**** ** ** ***, the ******* **** *****:

** ******* ***** ***** be ******* **** *** material ****** *** ******** is ********* *** ***********. No ********* **** ********* or ******* *** ***** display *** ********** ** a ******* ***** ***** be ******** ** *** rear ** *** ***** vehicle ******** ** ** registered ** *** *****.

Agree
Disagree
Informative: 1
Unhelpful
Funny

*'** ***** ** ** Derek ** ****** ********* on *******, *** * only ******** ***** *** they're ******* ** **, as ****, **** ***** ones. * *** * clear *** ** ** wife's *** ** *** and **** **** ***** off ** ** *** she *** ****** **** and **** ** **** it ***.

Agree: 1
Disagree
Informative
Unhelpful
Funny

**** ** *** ***** thing **** **** ** that ** ** ** easy ** ******. ** the **********, ***** *** lot **** ****** ****** just **** **** ****. Add **** ***** ** LEDs ** *** ***** frame ** ******* ********. Plate ****** *** ******* but **** *** ** illegal.

IPVM Image

Agree
Disagree
Informative: 1
Unhelpful
Funny

*****: *** ****** ** specifically ******** *** ********** to ******* **** ****** of ********, ***** ** consistent ****** ******. **** will ******* ***** ****** but *** *** ******** to ******* ** *** same ****/********.

****: *** ******* ******** Flock's ********, **** ********* poor ******** ** ***** plates, **** **** ******* visible ****** *** ***.

* **** *** ***** part ** *****'* ******* as '***** ***** *** a **** *****, *** not *** ****** **** front ******, ** ** designed ** ** **** back ******. ***** ***** complete *****.

**** *****'* **** ***** to ** ** ***, based ** **** *******, does ***** ******* ** badly (*************) ** ******* front ****** - ** both ***** *** **** plates *** *********?

Agree
Disagree
Informative
Unhelpful
Funny
Agree
Disagree
Informative
Unhelpful
Funny

** *** ******* ** bid ***** *** ***** of ****** ******* ** be ***** ****** *** exploring ***** ********* **** as ******. & ******** Does ******* **** ** there ** ********* ******* for ***** ************ ** have *** ***** *********** metrics?

*** ***** ******* * reached *** ** **** that * *** ********* operational ******** ***** ** in *** ************ ** $4500.

*** ******** ** ****** would ** ******* ***********.

***** ***,

Agree
Disagree
Informative
Unhelpful
Funny

*** ****** **** * look ** ******* ******* called *****.

Agree
Disagree
Informative
Unhelpful
Funny

**** ****** **** *** mention *** ************ ******* in *** **** *****. Was ***/**** ******* **** ? **** *** ****** cache *** ***** ******/******? Or, ****** ********* ** the ***** *** ** required ?

****, ** *** ***** for *** ***** ***** was * **** ********* antenna **** *** ******** connectivity, *** *** **** noted "*** ** ***". Is **** **** ** LTE **********?

Agree
Disagree
Informative
Unhelpful
Funny

********* *****, ***** *** for *** *********!

*** ************** **** ****, not ****. ****** ** fact, ***** ** ** WiFi ************ **** *** Flock ****** ******. ***** did *** ****/***/**** ****?

**** ******* ** ****** cache/direct ******, *** ****** uploads ****** ***********, ** long ** ***** ** a **********. ** *** connection ** *********/*******, ***** said:

*** ****** ** **** of ******* ******* ** how **** *** **** the ****** ** ** is. ** * ************ road ** *** ** up ** ** **** vs. *-* **** *** a **** ******* ****. The ******* **** ***** to ****** ***** **** first *** ******** ** capture *** ****.

*** ******* **** ** for ** *** **********.

Agree
Disagree
Informative
Unhelpful
Funny

>> *** ************** **** used,

**** ** *** ****** requirements *** *** ****** ? **** ******** ** upload ****** ?

>> ***** ** ** WiFi ************ **** *** Flock ****** ******. ***** did *** >>****/***/**** ****?

** *** ******** ******** :

IPVM Image

***** ****** ******** ********

***** *** * ********* to "******** ****" *** noted ** *** *** used. **** ** **** about?

****, ** ***** *** information ** *** ****** teardown ? **** *** vendor? ****** ********* ** the ****** ?

Agree
Disagree
Informative
Unhelpful
Funny

***** ** **** ************. It ** **** ****** set ** *** **** shut ***.

Agree
Disagree
Informative
Unhelpful
Funny

**** ****** **** ** these ******* *** **** compliant? * ***** ***** this ** ****** ** issue ** **** *** selling ** ********** ******** and ****** *********** ******* being *********.

Agree
Disagree
Informative
Unhelpful
Funny

***** ******* ****** ***** products *** **** *********. We *** *** ** a ******** *** ** don't **** *** ****** to ******* **** ** not ****.

Agree
Disagree
Informative: 1
Unhelpful
Funny

**** ***** **** ***** own *******?

*** ** ***, *** does?

Agree
Disagree
Informative
Unhelpful
Funny

** *** ***** ** this ******, ** ******* that *** ***** ****** is **** * ****** 5MP ****** ***** ******** both ***** ******* *** context. ** **** *******?

** **, *** *** analytics *** **** ** the ****** ("*****")?

* ******* ***** **** test: **'* *** ***** to ****** *** ******** the ****** *** ****, it's ******* ** *** how **** ****** ** never ******** ** ***. This *** ****** **** a ****** ** *** weeks, *** ***** *** analysis ** *** **** it ******? **** **** seemed * ****** **** compared ** ***** *** tests **'** **** ****.

* **** ******** ******* software *** *******, *** I've ***** ***** *** that * ***** ******** reliable *** *** ***********. I ******* *** **** thing ***** ***, ** Flocks ********* **, **** you **** ******, ** one ** *** *****.

Agree: 1
Disagree
Informative: 1
Unhelpful
Funny: 1

***, *** ****** **** one ****** *** ******** and *******. ***** *****'* disclose ******* ***** *** analytics *** ********* *** mentioned ** **** *********** with **** **** *** runs ** *** ******. If * *** ** guess [** ******* ****], I ***** *** **** the ****** ** ********* metadata *** ******* ** to *** ***** *** full **********. ****'* *** a *** ** ******** analytics ********, *.*.,********'* ********** ******* ********, ************* ************** * ****** **** a *** *** **********.

** ****'* ******* *** a **** *** ***** to *** ** ** missed *** ********, *** we *** **** ******* a ***'* ***** ** video. ** **** ******* mounted *** ***** ******** elsewhere *** ** ******** that ** * *** it *** *** **** any ******** *******.

***, **** *******: *****, our ****** ** *** extremely **** *******. *'* have ** ** **** and **** ** *** how **** ******** ******. It *** ******** ***+, but *** ******** **** what * **** ************ might ***. ******, ****** are ********** ***. **'* a ***** ***** ***** zone, ** ****** *** generally +/- ***** ** that. *****, ****** ***** are ****** ** *** faster ******. **** ** those ****** *** ****** to ******** ******* ** misses, *** ** *** not *** ***.

Agree
Disagree
Informative: 1
Unhelpful
Funny
Read this IPVM report for free.

This article is part of IPVM's 7,333 reports and 971 tests and is only available to subscribers. To get a one-time preview of our work, enter your work email to access the full article.

Already a subscriber? Login here | Join now
Loading Related Reports