Robotic Driving and More at Cornell

The researchers also will work with self-driving Segway transporters, which will be programmed to work together as teams for mapping and search-and-rescue-applications.

Physorg reports that Cornell is geting a new $1,473,121, three-year grant from the National Science Foundation to advance the self driving car.

“We’ll be looking at the next level of intelligence in the vehicle,” said Mark Campbell, associate professor of mechanical and aerospace engineering and the lead researcher for the project.

Skynet, Cornell’s entry in the DARPA Urban Challenge, drove itself through 55 miles of simulated city driving and was one of the top self driving cars.

Cornell Self Driving Approach

All of Skynet’s core code was open-sourced on Google Code in 2008. This includes our Artificial Intelligence, Path Planner, Target Tracking and Position Estimation algorithms.

Team Cornell divides its technical approach along five major sub-problems identified in the DARPA Urban Challenge: building a drive-by-wire vehicle platform, determining the location and orientation of the platform, sensing and tracking objects in the platform’s environment, estimating and extracting the structure of that environment, and planning and executing missions intelligently within that structure. These dividing lines allow solutions to these sub-problems to be developed in parallel, with rapid and seamless integration into a final prototype vehicle.

Cooperative Tracking
Campbell’s research group is working on cooperative tracking using Multiple UAV’s

Three ScanEagle UAV’s cooperatively tracking a target moving on a road

Sensor Fusion

Fusion theory applied to the Cornell vehicle in the DARPA Grand Challenge. Multiple sensors are quickly fused probabilistically to create a map for driving.