From a distance, this looks like one of the new autonomous cars. But for those brave enough to approach the window, they could see a man disguised as a driver's seat.

Why is someone driving around Virginia disguised as upholstery? For science, of course! The Virginia Tech Transportation Institute has confirmed that this is part of their experiment into how people interact with driverless vehicles, though the Institute has refused to comment any more on the issue. The Institute's website describes the study as a way to observe how humans react to autonomous cars and determine whether car companies should make any changes to their designs before unleashing driverless cars on the road. Not only is someone hidden inside the Virginia Tech van disguised as a car seat, but he sat completely motionless when passersby approached the vehicle and tried to communicate with him. This suggests that the study is likely focusing on how drivers interact with cars when there is no one in the driver's seat to communicate with.
While turn signals are the de facto way that drivers can predict another car's movements, many people don't realize how big of a role a driver's hand gestures or head movements play on the road. Sitting at a stop sign? It is common for drivers to motion to other cars or people to "go ahead" and cross the street first, and people throw up a "thank you" to let them know you're taking their offer. On the highway driving right next to a car? If you can't see a car's tail lights, seeing the other driver's head turn towards you can be an indicator that he or she is looking to shift lanes, allowing you to speed up or slow down accordingly. The Virginia Tech study is likely trying to measure these interactions. Without a driver to read, it becomes a lot more difficult for other drivers to predict what autonomous cars are going to do. This issue hasn't been sufficiently studied because when autonomous cars hit the road, there is usually an engineer in the driver or passenger seat monitoring the vehicle's movements. As a result, there isn't a lot of data on how the public responds to cars that have no one in them. Two years ago, a team of students from Stanford published a paper titled "Ghost Driver: A Field Study Investigating the Interaction Between Pedestrians and Driverless Vehicles." This study also featured a driver hidden inside of the car's seat. The goal of the study was to measure how pedestrians interacted with autonomous cars when crossing a crosswalk. Without a driver to gesture them across, pedestrians reported being nervous when the "driverless" vehicle inched too close to the crosswalk.
That study's findings led car manufacturers to consider implementing new technology to inform pedestrians that the autonomous vehicle recognized them and that it is safe for them to cross. When Mercedes-Benz unveiled their F 015 autonomous concept car, the vehicle projected a crosswalk onto the pavement to let pedestrians know it is safe to cross. Autonomous cars are programmed to identify lines on the road, signage, and obstacles. Not all crosswalks are accompanied by stop signs or traffic lights and few would want to purchase a car that stops at every crosswalk even when no one was in it. For this reason, autonomous vehicles will need to be taught how to interact with crosswalks, especially when there are people walking in it on the other side of the road or standing on the curb waiting to stop. Since many states have laws requiring cars to yield to pedestrians in crosswalks, this artificial intelligence will need to be expanded in order for the vehicles to be legal. As funny as it is to see someone hiding inside of a car's seat, the research is very important. Some estimate that there will be 10 million driverless cars on the roads by the year 2020. While much of the focus is on how the vehicles will interact with their environments, not enough is being done to see how the public interacts with driverless cars.

The State of Virginia wants your opinion: Should marijuana be legal?