We heard from Dr. Eran Ben-Elia about the work his team is doing to prepare for autonomous vehicles by researching with game simulations instead of traditional surveys.
Autonomous vehicles will have far-reaching effects on society, but it's the unintended consequences of these kinds of seismic changes that are always the most difficult to understand. One team that's taking a unique approach to researching and predicting some of these changes is the Game-Based Models Lab at Ben-Gurion University of the Negev. We caught up with the head of the lab, Dr. Eran Ben-Elia, to hear about his team's unique approach.
You can watch the video interview above or read the transcript below.
Ben-Elia said, "Surveys are very strong tools, but the problem is that you give responses that you have no responsibility for and you don't get any experience out of them. So, a lot of the phenomenon that we're dealing in transportation are experience-based and we need to understand how you react to situations in order to learn how you behave, and from how you behave, how to design and program technological solutions.
"It's nice to go into the real world, but the real world is very hard to manipulate. There are so many factors that I cannot control, so I prefer to use the lab. And in the lab, what we can do is we can play games. The games simplify reality, but keep the basic rules and the dilemmas in tact. And in this way, I can control factors that I can not control in the real world, and then design games where people reveal their behavior, they receive experiences, they react to different kinds of information that we provide to them in the game. They get a score, which they always react to. They can compete against other individuals. So, actually, we have a microcosm of what's really happening on the roads or in the transportation system in general, but in a much more controlled scenario, I would say."
Ben-Elia added, "Our studies are relatively cheap because we don't need to manipulate the real world. We're not putting expensive sensors. A lot of the things that are expensive is mostly hardware stuff, so we don't need to do any of these manipulations. We only work with software. Basically the only investment is software and a little bit on bringing volunteers into the lab to participate in the games.
"So, it's a university, we have students, students always need money. We have a roster that arranges these things and we publish the experiments and we say, 'This is the slot we have' and we ask for volunteers. We fill the slot and basically that's it. We [also] offer incentives. We always offer participation time and usually we also have [extra incentives] from their performances in the scores."
He concluded, "Because we are dealing a lot with social dilemmas in the transport sector and we have the technology that's interacting with that, we want to see things. For example, learn whether we can reroute people in a way that minimizes traffic congestion. We want to see whether people are willing to share rides on autonomous vehicles in order that we don't get stuck in traffic in autonomous vehicles that will be very safe, but will not be moving a lot because of traffic problems. And these are kind of complex systems where the games or the kind of simulations that we play with allow us to actually find what kinds of incentives people actually have to change their behaviors."
- Our autonomous future: How driverless cars will be the first robots we learn to trust (TechRepublic)
- US DOT announces autonomous car summit to speed development of self-driving tech (TechRepublic)
- Robots could write your next speeding ticket, if Ford's autonomous cop car comes to pass (TechRepublic)
- Ford sees autonomous vehicles as key cog of smart city (ZDNet)
- IT leader's guide to the future of autonomous vehicles (Tech Pro Research)