A driver avoids colliding with another vehicle. A bird lands on a tree branch. An astronaut docks a spacecraft. These different actions depend on the same skill: the ability to “brake” at the proper force at just the right time.

Armed with a three-year, $295,469 grant from the National Science Foundation, Brett Fajen is researching how most people are able to effectively control vehicles through braking and avoiding collisions.

“Braking appears to be a simple task. Even the simplest creatures are remarkably good at adapting to the immediate environment so they don’t systematically crash into the objects of their surroundings,” says Fajen, assistant professor of cognitive science. “Yet, no one has been able to build robots that can adapt to their environments like humans and other animals.”

The research could be used to improve transportation safety and collision, such as warning systems for vehicles and aircraft.

Such research also could be incorporated in designing mobile robots and increasing performance in the teleoperation of vehicles from remote locations, as well as lead to a deeper understanding of visual-motor impairment in humans.

To conduct his research, Fajen uses a driving simulator. Holding a joystick as a brake, participants sit in front of a large projection screen and look at a 3-D image of a moving road with an approaching stop sign. The task is simple: Apply brake pressure to avoid “crashing” into the sign.

Fajen can vary the task, such as making it appear that the driver is going faster. He also can change various textures and objects in and around the road, such as inserting white lines or making the road gravelly or smooth, hilly or flat. Fajen then compares and observes how people adapt to the varying changes in the surrounding as they brake.

The simulator records information 60 times a second, such as how fast and at what force one pushes on the joystick. It also keeps track of the virtual distance and speed of the approaching stop sign.

Published December 1, 2003