In 2007, I was part of the Honeywell team that contributed to the historic DARPA Urban Challenge, working alongside Carnegie Mellon University (CMU). The competition took place on November 3, 2007, at the former George Air Force Base in Victorville, California. Autonomous vehicles had to navigate a 60-mile urban course while obeying traffic laws and interacting with other vehicles.
My Role and Contributions
- Sensor Integration System: Designed and integrated multiple sensor systems to provide environmental awareness for the autonomous vehicle, coordinating data from LiDAR, cameras, radar and GPS.
- Perception System: Developed perception algorithms to process sensor data and detect and classify objects, including vehicles, pedestrians, road boundaries and traffic signs, in real time.
- Occupancy Grid Matrix: Created and maintained occupancy grid representations of the environment, providing probabilistic maps of obstacle locations to support path planning and collision avoidance.
The Challenge
The DARPA Urban Challenge required vehicles to handle complex urban driving scenarios, including:
- Obeying California traffic laws
- Navigating intersections with stop signs and traffic lights
- Merging into moving traffic
- Avoiding static and dynamic obstacles
- Passing other vehicles safely
- Parking in designated areas
The competition attracted 11 finalists, including teams from CMU (Tartan Racing, which won first place), Stanford, MIT and Cornell, as well as major corporations such as General Motors, Volkswagen and Honeywell.
Reflections to Come
This article will be expanded with personal reflections and lessons from the experience.
Leave a Reply