Over on Twitter, user @anelki has asked my thoughts on Clarendon’s new “Safety and Innovation Zone” pilot program:
For a couple of different reasons, I’m cautiously in favor of using automated smart cameras and other sensors in Arlington, including the new “Safety and Innovation Zone” pilot program on a block of Clarendon.
Although I’m hesitant of the purpose of detecting large crowds because that could be used to harass legitimate protests, it ultimately falls on our local leadership to react appropriately to large gatherings, not on the detection technology itself. The other stated goals of the specific “Safety and Innovation Zone” program are very good and have a lot of potential: listening for explosions and gunshots, and potentially detecting fire and smoke. The fact that this system does not have any video or photo capabilities on a fundamental technical level eases most of my concerns.
One use of the aspects as-is that I’m very much against is to detect verbal cries for help from individuals. Although it may mean well and could help public safety, having voice recognition in any capacity is a feature that would be ripe for abuse. One thing we should all have learned by now: if the government – even a nice local government – has the technical ability to do something bad, it eventually will do so. I would be fine with this one-year pilot program proceeding in order to test cybersecurity setups and refine requirements, but that feature would have to be removed entirely at the end of the trial period and before it is deployed elsewhere in the county.
Going beyond the “Safety and Innovation Zone” pilot program specific to Clarendon over the next year, I am also a cautious supporter of other smart camera and sensors programs.
A careful deployment of smart cameras can greatly help in our police reform efforts by allowing local law enforcement to spend their time and resources on fighting actual violent crimes in our community instead of minor traffic violations such as speeding and running red lights. Not that traffic issues are not important, but merely that automated enforcement is far more effective and efficient at catching them than having police constantly patrolling intersections. A good automated system can easily catch offenders 24/7/365 without creating police-civilian encounters that might end up violating civil liberties or turn dangerous for either side.
There has also been a rash of reported noise complaints from loud vehicles around Clarendon and other parts of the county. A well-designed system can do a much better job of identifying these noisy vehicles throughout the day and night than posted law enforcement officers could. From there these noisy offenders can be warned, educated, or fined as appropriate using information from their vehicle registration.
Smart sensors could also be used to easily give traffic signal priority to busses, cyclists, and pedestrians – speeding up our mass transit network and reducing the chances of injuries.
As these automated systems are developed, it is critical that they be very carefully designed and tested for cybersecurity, make sure they are isolated from other government agencies that could abuse the data, and never (under any circumstance) be enabled with facial or other biometric recognition features. This process would not be a fast or cheap one. We must not rush into it. I could easily see this build-out taking 5-10 years to properly design and deploy. Part of this proper deployment should involve lots of small, temporary pilot programs with limited scopes over the first few years.
I provided some of the above points to Sustainable Mobility for Arlington County (SusMo)’s 2021 Candidate Questionnaire, where you can look at Question #3 for the issue of automated traffic enforcement and public safety.