Tuesday, November 1, 2016

How to accelerate adoption of Connected/Autonomous Vehicle technology

Innovation will always outpace regulation (I use that phrase regularly, so bear with me). In some cases, the regulatory aspect is not a great concern but this is not so when it comes to connected/automated vehicles. As fascinating as the CV/AV space is, there are still some huge technical and non-technical hurdles that must be overcome.
  • Designing infrastructure and controls that work mixed environments with traditional, assisted, and autonomous vehicles; 
  • Imbuing "judgement" in the artificial intelligence of the AVs - How does the car decide which object to hit if it is unable to avoid a collision;
  • Securing the data collection and controls from hacks;
  • Defining liability issues when an autonomous vehicle is involved in a collision;
  • Communications standards between various manufacturers as well as the infrastructure.
In this article from the Texas A & M Transportation Institute (TTI), authors Jason Wagner and Ginger Goodin discuss the policy question of data security: Who owns the vehicle data and who gets to see/use it?

Personally, I am always interested in tracking my own data, typically associated with my workouts (read: bike rides).  Pulling similar data on vehicle health and driving statistics also intrigues me but I don't necessarily want to provide that data to anyone without my permission. 

I've been known to engage in spirited driving when no one is around.  Sometimes I speed down the tollway, hitting booths faster than someone going the speed limit would.  My maxim is that I am happy to share my data, just don't use it against me.  Would my insurance company raise my rates due to the occasional rapid acceleration/deceleration or high-G turns, even though I have never caused an accident (true statement)?  Could the police retroactively ticket me after mining my information from some pool where my data ended up without my knowledge?

My response to these types of scenarios, and I'm not unique, would be to turn off the data, if possible.  The problem with me not sharing my data, though, is that it deprives the "good" analysts of another sample and, as any statistician will tell you, the more samples, the better the understanding.

As the Internet of Things expands into every aspect of life, these questions must be answered in a way that satisfies private entities' concerns within the public/social framework.  This use case just highlights some of the challenges we are experiencing in this space but is reflective of any new technology.

There is a regulatory drag that can slow down wholesale adoption of these types of technology, similar to what we see with the FDA and new drugs.  In this space though, I have found that it is not a Luddite reaction on the part of regulators but a lack of understanding about the technology and a lack of time to think through the various implications of deploying it within the populace.

Something with as grand a scope and scale as connected/autonomous vehicles creates a prime opportunity to involve the regulators and legislators much earlier in the cycle.  By doing so, the auto manufacturers can help drive adoption, but it will require a new level of cooperation among them, even as they try to gain competitive advantages.  It also helps to have academic researchers involved on the development side, driving standardization and a more objective viewpoint on which the regulators can rely.

That's why I'm glad to be involved with an organization like TTI which is bringing together all of these parties to work through the regulatory questions as the technology is developed, thereby accelerating the adoption so we can realize its benefits sooner and, hopefully, allowing me to collect my data without my insurance rates going up.

No comments:

Post a Comment