Urban Challenge



Description



Testing early and often was a key ingredient in Tartan Racing's
recipe for Urban Challenge success.


Tartan Racing took a multi-pronged approach to the daunting challenge of navigating the dynamic environment of a city:

Organize and orchestrate concurrent software components to determine sequences of tasks, process sensor data, and control the vehicle. Response times had to be less than a second!  Other software components continuously monitored the status of individual tasks to find out whether they were
successfully completed.




Sense, differentiate and localize objects (such as buildings and cars) and environmental features (such as lane markings, curbs, and sidewalks), both fixed and moving.  The vehicle used radar, ladar, and video sensors to perceive its environment and GPS and IMUs to establish its position.




Control actuators to ensure safe and efficient driving in intersections, parking lots, traffic circles, and similar city features.




Plan and replan the most efficient routes through a network of streets, taking into account constantly-changing conditions.




Retrofit two stock Chevy Tahoe SUVs to enable computer control of their steering, speed and gears.




Achieve robustness by applying well-known principles of systems engineering and testing, using both simulation and live tests.  The goal was to make the vehicles as reliable as possible for the race.

NREC faculty and staff took on key leadership roles in conquering these technical challenges.

Bryan Salesky was the software lead for Tartan Racing and was responsible for managing requirements definition, architecture, design, implementation and testing.  He brought NREC’s CMM Level 3 process management expertise to the software development effort.




Tony Stentz adapted Field D* planning algorithms to the Urban Challenge application.

Al Kelly adapted Ranger motion planning algorithms to enable driving behaviors appropriate for urban settings.

By participating in the Urban Challenge, NREC broadened its autonomous navigation capabilities from off-road, complex terrain to urban and highway environments.