What The Run Over Child Means For The Future of Security Robots

By: Brian Karas, Published on Jul 21, 2016

Robot security guards recently received unprecedented news coverage. Unfortunately, not good:

We spoke with Knightscope CEO William Li [link no longer available] about this incident, and what it means for the robot guard market in general.

In this report, we analyze the current state of robots as guard alternatives, what their limitations mean for today and future projections for security robots.

***** ******** ****** ******** received ************* **** ********. *************, not ****:

** ***** ****************** ******* ** [**** no ****** *********] ***** this ********, *** **** it ***** *** *** robot ***** ****** ** general.

** **** ******, ** analyze *** ******* ***** of ****** ** ***** alternatives, **** ***** *********** mean *** ***** *** ****** projections *** ******** ******.

[***************]

Stanford **** ********

* *********** ** **** on ****** ** *** Stanford **** ** ********, CA *** ****** ** avoid * ***** ***** and ***** ** ********* with *** *****, *** allegedly running **** ***. ** case *** ****** **, here ** * *******:

*********** ****** ****** *******, ************ **** and ***** ****** ******* incident, *** ********* ****** that *** ***** *** over *** ***** (*** not **** **** ********), stating

**** *** ********* ********* the ***** *** *** path *** *****, *** machine ******* **********

*** **** *** *** robot ******* **** *** child, ** ****** *** understanding ** *********** ***** what ** ** ** that ********* (****, ***** on ****** ** *** child, **** *** ****, etc.)

Security ******** **** ****'* ***

***********'* *** ******** *** current ***** ** ******** robots ** ****'* ***, stating **** ** **** at * ***** ***** technology ** ***** ** be **************, *** ** a ***** ** *********** a ******** ** **** it **** ********** ******. Unlike ***, ***** **** roughly ** ***** ** really ****** **********, ** expects *** **** * years ** **** ******* advancements ** ********** **** enables ****** ** ****** practical ** ****** ******.

** *** ********* ***** of ** ********* **** of *** ********** ******* of *** ******* *********** robots: 

******* *** ***** ** sensors *** ************, *** device ******* ** ** somewhat ****** "*******", ***** in *** *********, ***, according ** *********** "********* ********* ** ************* 1 ***":

*** ***** ********, *** the ***** *****, ********* how *********** ***** ********** still **. ***** ****** **** run **** **** ***** and ***** ****** *********, they ** ***** ****** the ******* ** ****** the ********* ****** ********** further, ** ***** ** apology. *** ****** *** forced ** ****** ** low ***** ** ***** to ********** *** ***** inability ** ******** ******** in * ***** ********** by ******. * *** pound ******, *** ***** aware ** *** ************, traveling ** ***** ***** pose * ****** ******.

Machine ******** **** ***** ********

**** ***** ***** ******** technology ************ **** ***** impact ****** *** ********, machine ******** *** *** CEO's ***** ******.  ** do ***** **** ******* learning/neural ******* ********** *** been ******* ************ ** recent *****.  *** *********/****** *********** ***** ** ******* ** machine ******** ********** ************ ** security ************. 

Mobility ****** ********* ** ********

* ***** *****, *** liability, *** ******* ****** is ********, ** ** clearly ********* ** **** a ******** **** ** both ******* ** ****** about, *** ***** ** without ********* ****** ** the ****. ****** *** batteries *** ***** *****, which ****** ******* ********, and *** **** *** device **** ********* ** it **** *** ** to * *****, ** mis-navigate ** **** *** fall *** * **** or *****.

******** **** **** ****, for *** ***** ** a *********** *****, * user *** ****** ******* fixed ******* ***/** ****, giving **** ******* ******* to ******* *****, ******* having ** ***** ***** mobility ** ********* ******.  

Robots ** ***** ************ *******

*** **** ******* ******** technology ********* ** ********** robotics *** ** ******* to ***** *******, ***** also **** *** ******* of ********** ******* ** an ****, *** *** ******* to **** ** ******** or ***** ***** ******** to **** **** *** a ***** ** ****.

******** ****** *** ***** positioned ** ** *********** to * ***** *****, but **** *** **** ** * ******-**-*-****** base. **** ** *** cost *** ********** ** the ****** **** **** the ****** *** ******* required ** **** ** both ****** *** ********-*****, costs **** ***** ** simply **-********* ** ****** more ******* ** * site.

Knightscope's ********* ******* ***** *** ******

***** ***** **** ** would **** *** ****** to ****** ** *********** as *****'* **** ****, Knightscope *** **** "*** *** we **** * ****, or ******* * ********" would ** * ******* point *** ********, ******** making ** * "********* not ** **** *** technology" *** ********* **** the ******** ****. ***** this ***** ** * good ******** *** ********* more ***** ******, ** also ***** **** * goal **** **** **** more **** * ***** to *****. *** ******* ** a ***** "****** * life" ******** ***** ***-*******, though *** ***** *** be **** ** ******* an ***** ** ** event, ** **** *** same *** *** ****** of ***** ******** ******* do *****.

Equivalent ** ****'* ***** ********

***** ***** **'* **** expensive, ***********, *** ********** simplistic ** **** **** could **, **** *** offer **** ******** ** a ***** ***** ** users **** **** ******* and * **** *********** to *** (*.*., *******, science, **********).

*** *** *** ***** adopters ** ******** ****** that *** *** *** high ***** (*********** ******* $50k *** **** *** robot) *** *** ****** the *********** (* ***, near **** ****** *****)? * mall ******** ***** *********** does *** **** ** be *** *****. ** surely ** ** ************ novelty *** *** **** of * ******** *****.

***********'* *** ********* *** they *** ******** ** rolling *** *** ********** in *** *** **** first, *** **** ** slowly ****** ****** *** US *** **** ****** deployments.  *** **** ********* are ***** ** ********** early ********, **** *** benfit ** * **** cooperative *******, ***** **** there *** *** *********** indicators ** ***** ***** viability.  **** *********** *** proven *********** ** "******" ****** sectors, **** * ***** of ********** ** ******** Detroit, ** * ***** mall ** ***** ** will ** * **** larger ********* ** ** emerging ****** *** ***** guards.

[UPDATE ****** *, ****]

*********** *** ****** **** ***** *******, ********** * statement ** ***** ******* release***** *** ***** ******* backwards. *** ******* *** states *** ***** *** running ********, ******* *** unit, *** ******* ****. However, *** ****** ** which *** ***** *** running ****** **** *** change *** ******* ****** surrounding *** *********.

Comments (24)

Lets go find one and barricade it in an area like a roomba lol

If these things become even mildly popular I would expect bored kids to come up with various games or ways to troll the units, similar to what you described.

I can envision a K5 dutifully patrolling a parking lot with a "Kick Me" sign on its back.

Robot vendors don't seem to have thought about edge cases. Which is pretty scary considering some of them are selling into nuclear power stations. This is one of those cases where "what's your security story" should produce an answer other than "huh?".

A legitimate question... What is the better approach (better being defined as: if you were the company trying to get these to market and raise more money), should they spend years in the lab and try to test and design for every possible edge case, or should they make it "safe enough", set it loose, and then iterate on the feedback?

Taking the Silicon Valley approach of "fail early, fail often" to a 300lb stack of batteries and motors really changes the considerations a bit.

Yo, yo, yo, security robot.

Where is your video??

They ordered body cameras but Taser was a no bidder.

Where is your video??

We asked for the video from the robot. It is not being provided. My guess is that it would not make things any better and might make things far worse seeing a 1 year old colliding / being run over by a 300 pound robot.

Kit, the car traveled at much higher speeds and never ran over a kid. Where is the Hoff when you need him.

My guess is that it would not make things any better and might make things far worse seeing a 1 year old colliding / being run over by a 300 pound robot.

Even if 'it would not make things any better', it would be better to be transparent. If it would 'make things far worse' then I would have to think that this statement is inaccurate:

a child left the vicinity of his guardians and began running towards the machine. The machine veered to the left to avoid the child, but the child ran backwards directly into the front quarter of the machine, at which point the machine stopped and the child fell on the ground.

'Running backwards' is tricky, try it. Best way to do it is to run forward and have someone run faster behind you with a camera... :(

it would be better to be transparent. If it would 'make things far worse' then I would have to think that this statement is inaccurate

I am quite skeptical of that statement for the reasons you bold above and its generally convoluted / unclear manner in which it was written.

So, yes, my suspicion unless this goes to court they will do well not to show it to anyone. Most people do not like to see 1 year olds injured by anyone, including robots...

Asked about what it would take for robots to become as commonplace as today's mall cops, Knightscope CEO said "The day we save a life, or prevent a casualty" would be a tipping point for robotics, possibly making it a "liability not to have the technology" for customers like the Stanford Mall.

This is a really bad answer.

Automation for the sake of automation is a gimmick. Automation needs to be demonstrably better than human alternatives in some, even multiple aspects to reach a 'tipping point'.

In every example of automation reaching a 'tipping point', it is because the cost of technology is cheaper than human resources, automation does a superior job to a human, or automation can withstand environments that humans are harmed by.

For example: automobile assembly robotics. It wasn't until robots could lift more, quicker, more accurately, for longer hours, and in an hot, fume filled environment that car companies replaced human 'panel beaters'.

Knightscope's CEO says saving lives or preventing casualties is something his current robots are unable to achieve. That's fair, but then again why would anyone be interested in buying something that just rolls around? How hard is it to sell that!?!

Similarly, the Robolliance says that robots can be used for jobs that are dirty, dull, or dangerous, but in security these are almost always in environments that make it hard for a robotic platform to be practical.

The automotive robot comparison is interesting, I was thinking about those and trying to draw comparisons when writing this article. In that case those robots have even less intelligence than the K5. They are fixed-location units, designed to repeat a very specific pattern endlessly, without factoring in much of their environment at all (as a side-story, I once watched a robot in a glass manufacturing facility that was programmed to place the metal "knob" for a rear-view mirror on windshields as they moved down an assembly line happily smash the knobs through about 50 windshields in a row, as it had no idea its parameters were off).

Automotive robots replaced humans in what were extremely boring and precisely repetitive jobs, and even then they are far from perfect. Scaling that up to a mobile unit that can interpret ever-changing surroundings is no small task, even when you factor in technological advancements.

Alot of the youtube pranksters live in california, lets put one of them up to messing with the robots

The robot has been suspended without pay while the investigation is pending.

Newer models will not need to roll over citizens, relying on an electrically controlled riot stick, when necessary:

The issue is the technology and analytics are not advanced enough to deal with a chaos situation.

You can assume an adult walks and moves in a way that is defined in a mail.

A kid doesn't, they often run off and find a robot more toy like so will test it. It's how children learn they test things. Everything we do is built around these invisible laws except one thing. Cars, guess what kids walk out in front and die.

I think it's a bit unfair on the technology, if it was at commercial site closed off patrol that would make sense. I wouldn't put it in a mail with kids doing random things.

I wouldn't put it in a mail with kids doing random things.

Then you evidently wouldn't make it as a Silicon Valley marketer! ;)

I agree with you operationally and technologically. I still do not understand what real security role that robot could play at Stanford mall compared to its cost and functionalities.

But it (was) great marketing for Knightscope and probably for the mall as well until this incident. Then, of course, disaster.

But having your product as a walking billboard across the street from the Sand Hill Road venture capitalists? That's tough for a startup CEO to pass up.

"I still do not understand what real security role that robot could play at Stanford mall compared to its cost and functionalities"

Security guards and electric are the 2 largest expanses for shopping centers after they are built. Management is always trying to find ways to lower these bills. The goal was not to increase security but more as a test to cut expanses.

Security guards and electric are the 2 largest expanses for shopping centers after they are built.

Citation for that?

Even assuming this is true, at $50,000 per year for this robot, how much are they really going to save? And a mall security guard can actually do things, like give directions to people, mediate basic disputes, etc.

Hi John, I know this from direct conversations with top property management companies.

I agree with you on cost but the plan is according to the article below is to have a combined machine/Man guard unit and the technology will get more cost effective.

Universal Protection Service the current security guard providor for stanford mall have partnered up with knightscope to promote "Machine a service" and this was a test sites.

http://www.securitysystemsnews.com/taxonomy/term/10451

I don't believe this Robot will make it to south CA in September after this issue.

I read the article but I am no clearer about what the value proposition will be. What will the robots be doing that will provide value? Is the belief that robot will deter crime or?

Related, can anyone provide any clarity what the robots were doing for security at the Stanford mall?

Btw, as for largest expenses, security does not appear to be in the top 2, e.g.:

Services, from what sources say, includes security, landscaping, trash removal, etc.

45% for insurance is looking like a good deal for the Stanford mall at this point.

Hi John, Most of these large property management companies are self insured but I believe they are looking at expanses that they can control. Insurance/tax is not one them.

As for the value, I visited their website and could not find anything interesting other than “see, feel, hear and smell” enabling a unique approach to sensor fusion and analytics"

I won't be surprised if this was a free demo. Stanford mall is owned by Simon properties, the largest shopping center operator in the US.

Knightscope forgot to put a big bright warning sign on the robot "DO NOT TOUCH" lol

Read this IPVM report for free.

This article is part of IPVM's 6,438 reports, 865 tests and is only available to members. To get a one-time preview of our work, enter your work email to access the full article.

Already a member? Login here | Join now

Related Reports

Cobalt Robotics Presents Indoor Security and Access Robots on Jun 03, 2020
Cobalt Robotics presented indoor security robots at the May 2020 IPVM...
Coronavirus Shuts Down ADT Door Knockers on Mar 26, 2020
Coronavirus has another victim - this time, alarm giant ADT has stopped all...
Cobalt Robots Respond To Coronavirus on Apr 20, 2020
Startup Cobalt Robotics is responding to coronavirus with new features...
Startup Videoloft Presents Cloud Storage on May 27, 2020
Videoloft presented offsite cloud storage at the May 2020 IPVM Startups...
Latest London Police Facial Recognition Suffers Serious Issues on Feb 24, 2020
On February 20, IPVM visited another live face rec deployment by London...
Cisco Video Surveillance Is Dead, Long Live Cisco Meraki Video Surveillance on Feb 11, 2020
A dozen years ago much of the industry thought that Cisco was destined to...
Uniview Deep Learning Camera Tested on Jul 14, 2020
Uniview's intrusion analytics have performed poorly in our shootouts. Now,...
China Surveillance Vulnerabilities Being Used To Attack China, Says China on Apr 07, 2020
While China video surveillance vulnerabilities have been much debated in the...
Defendry Presents AI Active Shooter Security System on Jul 14, 2020
Defendry presented its Active Shooter security system at the May 2020 IPVM...
ADI Branch Burglary on Apr 03, 2020
A security systems distributor branch is an odd target for burglary but that...
Colombia's President Promotes Bad Hikvision Fever Camera Setup on Jun 17, 2020
Colombia's President Iván Duque has promoted a haphazard Hikvision fever...
Sunell Panda Cam Body Temperature Measurement Camera Tested on May 14, 2020
Sunell is far less well known than its gargantuan domestic competitors Dahua...
Anixter Runs Fake Coronavirus Marketing Using Shutterstock Watermarked Images on Jul 24, 2020
Coronavirus faked marketing is regrettably commonplace right now but Anixter...
Faulty Hikvision Cali Colombia Fever Camera Implementation on Jul 20, 2020
The mayor of one of Colombia's largest cities has promoted a faulty Hikvision...
Manufacturers And Reed Speak on Planned July 2020 ISC West on Mar 11, 2020
Last week, the drama of the March 2020 ISC West cancellation transfixed much...

Recent Reports

Dangerous Hikvision Fever Camera Showcased by Chilean City on Aug 07, 2020
Deploying a fever camera outdoors, in the rain, with no black body, is...
"Grand Slam" For Pelco's PE Firm, A Risk For Motorola on Aug 07, 2020
The word "Pelco" and "grand slam" have not been said together for many years....
FLIR Stock Falls, Admits 'Decelerating' Demand For Temperature Screening on Aug 07, 2020
Is the boom going to bust for temperature screening? FLIR disappointed...
VSaaS Will Hurt Integrators on Aug 06, 2020
VSaaS will hurt integrators, there is no question about that. How much...
Dogs For Coronavirus Screening Examined on Aug 06, 2020
While thermal temperature screening is the surveillance industry's most...
ADT Slides Back, Disappointing Results, Poor Commercial Performance on Aug 06, 2020
While ADT had an incredible start to the week, driven by the Google...
AHJ / Authority Having Jurisdiction Tutorial on Aug 06, 2020
One of the most powerful yet often underappreciated characters in all of the...
SIA Coaches Sellers on NDAA 889B Blacklist Workarounds on Aug 05, 2020
Last month SIA demanded that NDAA 899B "must be delayed". Now that they have...
ADI Returns To Growth, Back To 'Pre-COVID Levels' on Aug 05, 2020
While ADI was hit hard in April, with revenue declining 21%, the company's...
Exposing Fever Tablet Suppliers and 40+ Relabelers on Aug 05, 2020
IPVM has found 40+ USA and EU companies relabeling fever tablets designed,...
Indian Government Restricts PRC Manufacturers From Public Projects on Aug 04, 2020
In a move that mirrors the U.S. government’s ban on Dahua and Hikvision...
Directory of 201 "Fever" Camera Suppliers on Aug 04, 2020
This directory provides a list of "Fever" scanning thermal camera providers...
Face Masks Increase Face Recognition Errors Says NIST on Aug 04, 2020
COVID-19 has led to widespread facemask use, which as IPVM testing has shown...
Dahua Loses Australian Medical Device Approval on Aug 04, 2020
Dahua has cancelled its medical device registration after "discussions" with...
Google Invests in ADT, ADT Stock Soars on Aug 03, 2020
Google has announced a $450 million investment in the Florida-based security...