Pravin Varaiya doesn’t drive much; the 62-year-old UC Berkeley engineering professor usually walks to his office. And when he does need to drive, he keeps it simple—his personal vehicle is a 1965 Volvo.

Despite the low-tech approach to his own daily commute, Prof. Variaya’s work puts him on the cutting edge of transportation technology. From 1994 to 1997 he was the director of the California PATH research program, a collaboration between the California Department of Transportation, universities, and industry dedicated to solving California’s transportation problems. He continues to study transportation technology today.

Much of Prof. Variaya’s research sounds like something from a Jetsons cartoon. He spends his day thinking about things like how to deploy “smart dust” sensor networks that can autonomously track traffic and how to create “intelligent” traffic signals that automatically optimize traffic flow. BDTI asked Prof. Varaiya to share his perspectives on the outlook for autonomous, self-driving cars.

BDTI: What got you interested in autonomous vehicles?

Prof. Varaiya: The technical challenges were exciting, and the opportunity to test the technology in real vehicles—and not just develop algorithms—made this a fun project. I later discovered that the technology could dramatically increase highway capacity, which added a public policy appeal to the work.

BDTI: What about your funders—why are they interested in autonomous vehicles?

Prof. Varayia: When we compare projected traffic growth to the expected highway capacity growth, we see that demand will outstrip capacity for the foreseeable future. Part of the problem is the difficultly of building highways in urban areas. After the wave of urban renewal in the 1960s, the social and political opposition to new roads became intense. It became clear that adding highways only improves traffic for two to five years, and then you get congestion again. Moreover, building new highways breaks up communities; it forces you to demolish housing, relocate residents, and so on.

Mass transit is often proposed as a solution. But while where is a strong support for transit in principle, there isn’t the same enthusiasm for the taxes to support transit. And in low-density areas, mass transit isn’t effective at all.

Autonomous vehicle technology promises to provide a huge increase in capacity without building new roads and without the expensive infrastructure associated with mass transit.

BDTI: Are there any fully autonomous transportation systems in use today?

Prof. Varaiya: Yes, there are. Fully automated trains are currently used in a few locations, mainly in the U.S., Europe, and Japan. Most of these systems are used in secure locations such as airports—the automated train system at the Atlanta airport is one example of this. These trains operate in an environment with very limited access, so the safety issues are minimized. These systems are also relatively safe because they operate at very low speeds.

Although these trains are fairly slow and simple, they still present significant design challenges—the level of automation in these systems is quite high.

BDTI: What are the motives behind these systems?

Prof. Varaiya: Cost is one motive: these systems are less expensive to build and operate than a traditional rail system. An automated system is also more flexible. The infrastructure is lighter and less expensive than traditional rail infrastructure, so it is easier to change the system. And while traditional rail systems operate on fixed schedules, an automated system can respond to changes in passenger loads.

On the other hand, I haven’t seen the numbers to know if the economics justify the investment. It may be that these systems are built partly because of the novelty aspect.