University of Southampton (ECS) researchers recently won the inaugural Lemonade Game Tournament by defeating research teams from Princeton University, Brown University, Rutgers University, Carnegie Mellon University, and the University of Michigan. The teams had to create an agent that recognizes a collaborator in a repeated three-player zero-sum game. The agent must cooperate with the collaborator in order to defeat the third player. However, the agents are not able to communicate directly with one another, so they must indicate their willingness to collaborate by signaling. The ECS team designed an agent that classified the style of behavior that other agents in the game are playing, and used this information to rank their potential as collaborators. The researchers are working on improving the agent for the next tournament.

View Full Article

For More Information Visit: http://www.cpccci.com

Scientists at the Center for Automotive Research at Stanford (CARS) have redesigned an Audi TTS with computers and global positioning system (GPS) receivers so that the car can drive itself. The car will attempt to scale Pikes Peak without a driver at race speeds following a GPS trail from start to finish. “Our first goal is to go up Pikes Peak at speeds resembling race speeds, keep the car stable around corners, and have everything work the way we want it to,” says Stanford’s Chris Gerdes. The car has reached speeds of 130 miles per hour without a driver during test runs at the Bonneville Salt Flats in Utah. It uses differential GPS, which corrects for interference in the atmosphere, and can locate the car’s position on Earth within about two centimeters. The car measures its speed and acceleration with wheel-speed sensors and an accelerometer, and also employs gyroscopes to control equilibrium and direction. “The computer puts all this information together and then compares it to a digital map to figure out how close the car is to the path that we want it to take up Pikes Peak,” Gerdes says.

View Full Article

For More Information Visit: http://www.cpccci.com

Rutgers University researchers have developed an algorithm to help find open parking spaces, using ultrasonic sensors, global positioning system (GPS) receivers, and cellular data networks. The goal is to create Web-based maps or additions to navigation systems that make parking availability data accessible to those looking for a parking space. The researchers, led by Rutgers professors Marco Gruteser and Wade Trappe, say that data could help alleviate traffic congestion and cut down on energy use. To implement their system, the researchers mounted ultrasonic distance sensors on the passenger-side doors of three cars, which over two months collected parking data in a limited urban area. They then created an algorithm that converted the ultrasonic data into information on available parking spaces. Combining that data with GPS data, the algorithm generated maps that were more than 90 percent accurate. To distinguish parked cars from other objects, such as trees or recycle bins, they measure the length and width of each object using the ultrasonic sensor readings. Gruteser says the parking availability information could be distributed over the Internet and used by drivers to decide whether to park on the street or use a garage.

View Full Article

For More Information Visit: http://www.cpccci.com