By Ginger Goodin And Christopher Poe
Despite their rarity, reports of crashes involving self-driving cars underscore an important point: all of the players in industry, government and research must work to strike a balance between accelerating technology with the potential for positive outcomes and minimizing public and private risk.
Technology will surely solve some problems, but it may also create new ones.
Our current human-driver system is failing. We are simply killing too many people and accepting too many crashes each year. Automation could reduce or eliminate most human-driver errors that account for crashes; at the same time, early self-driving systems (i.e., sensors, software, interfaces) could result in errors and crashes that humans might avoid.
Automated cars that are safer than human drivers could reduce the frequency of crash-related traffic delays, but a proliferation of those cars could exacerbate more routine roadway congestion.
Self-driving technology may increase trip frequency and generate more economic activity, but without greater efficiency and electrification, more trips could produce more delay and tailpipe pollution, posing risks to mobility and human health.
Advancements sometimes involve unintended consequences, so the competition to commercialize automated vehicle technologies must be balanced with the need to be completely satisfied — through robust and thorough research and testing — that these technologies are as foolproof as possible before introducing them on public roadways.
And accounting for the endless possibilities of the driving environment, that’s going to require a lot of testing — in both physical settings and simulated, virtual environments.
Consider just a few of the potential variables: lighting conditions (daytime, nighttime, dusk and dawn); weather (ice, snow, and flooding); wildlife (deer and feral hogs); unique roadways (most states don’t have frontage roads or the Texas U-turn); and human-driven cars sharing the road space for many years.
Research on controlled test facilities is an essential first step to deployment of autonomous vehicles on public roads. The Texas Automated Vehicle Proving Grounds Partnership — involving the Texas A&M Transportation Institute, The University of Texas at Austin, and Southwest Research Institute — provides a comprehensive network of proving grounds and test tracks where vehicle automation can be tested for a wide range of scenarios to be expected in the real world.
The partnership also includes seven regions with roadways designated as technology deployment sites.
In the Fort Worth region, those include the I-30 corridor along with streets in Arlington around the entertainment district, and The University of Texas at Arlington campus. Real-world environments like those are vital to demonstrating the success of these technologies to the public.
Last year, the City of Arlington started a pilot project with automated vehicles, Milo by EasyMile, around the Ballpark in Arlington and the AT&T Stadium. This was the first place in Texas where the public could ride in a driverless vehicle. And, in July, self-driving startup Drive.ai will begin testing its own vehicles on limited routes in Frisco, Texas, to serve patrons visiting sporting venues, shopping and restaurants.
To better grasp the legal, public policy, and governance implications of self-driving cars, TTI researchers are working closely with the Texas A&M University School of Law in Fort Worth.
Beyond investment, governance for the deployment of self-driving technology is a natural extension of government’s role to ensure the safe and efficient operation of public roadways.
Companies like those leading the autonomous travel revolution move rapidly and nimbly. Government agencies typically do not. The challenge for government is to keep up with innovation — which may emerge so quickly that it’s in widespread use ahead of any efforts to create policy — or, in some cases, responsibly regulate it.
They’re called “disruptive” technologies for good reason.
We have seen that challenge with Uber, Lyft and other transportation network companies, as well as dockless bikes and scooters which emerged seemingly overnight.
However, government must seek to understand and project their impacts on congestion, public transit use and personal safety to create a policy environment that balances technology adoption and public safety.
Similarly, on the road to an automated driving future, there will be inevitable trade-offs along the way. With a purposeful approach to research and testing, we can manage those consequences in a way that minimizes public risks and accelerates positive outcomes.
Christopher Poe, Ph.D., P.E., is an Assistant Agency Director and Connected and Automated Transportation Strategy Lead, and Ginger Goodin, P.E., is a Senior Research Engineer, both at the Texas A&M Transportation Institute.
This article was originally published in the Star Telegram, July 20, 2018.