-1- 2008 Florida Conference on Recent Advances in Robotics, FCRAR 2008 Melbourne, Florida, May 8-9, 2008 Lessons Learned at the DARPA Urban Challenge Carl Crane, David Armstrong, Antonio Arroyo, Antoin Baker, Doug Dankel, Greg Garcia, Nicholas Johnson, Jaesang Lee, Shannon Ridgeway, Eric Schwartz, Eric Thorn, Steve Velat, and Ji Hyun Yoon University of Florida Gainesville, FL 32611 352-392-9461 ccrane@ufl.edu ABSTRACT This paper describes the intelligence components associated with the system design developed for Team Gator Nation’s submission to the 2007 DARPA Urban Challenge. In this event, vehicles had to navigate on city streets while obeying basic traffic laws. One of the major challenges was interacting with other vehicles such as at intersections. To address these challenges, a hybrid Toyota Highlander was automated and instrumented with pose estimation (GPS and inertial) and object detection (vision and ladar) sensors. A control architecture was developed which integrates planning, perception, decision making, and control elements. The intelligence element implements the Adaptive Planning Framework which was developed by researchers at the University of Florida. This framework provides a means for situation assessment, behavior mode evaluation, and behavior selection and execution. The paper describes this architecture and concludes with lessons learned from participation in the Urban Challenge event. Keywords autonomous ground vehicle navigation 1. INTRODUCTION In DARPA’s vision, “The Urban Challenge features autonomous ground vehicles maneuvering in a mock city environment, executing simulated military supply missions while merging into moving traffic, navigating traffic circles, negotiating busy intersections, and avoiding obstacles.” Moving the challenge into an urban setting adds structure and complexity to the Grand Challenge problem. Previous success relied on a single mode of operation, without interaction with the environment beyond simple traversal. Success in the Urban Challenge will require numerous modes of operation and complex interaction with the environment. It is expected that the urban environment will also hamper the use of GPS for localization, further complicating the challenge. The specific problem to be solved is detailed in the Urban Challenge Technical Evaluation Criteria document [1]. Here the problem is organized into four categories, i.e. Basic Navigation, Basic Traffic, Advanced Navigation, and Advanced Traffic, each of which is more complex than the previous. Upon reviewing this document, the authors identified the following set of technical challenges: 1. pavement (road) detection and lane detection 2. detection of static obstacles 3. detection and classification of dynamic objects 4. environment data representation and sensor integration 5. localization 6. reconciliation of differences in estimated global pose, a priori data, and sensed information 7. high level mission planning 8. determination of appropriate behavior mode 9. smooth transition of vehicle control between behavior modes 10. interprocess communication and coordination of multiple threads on multiple computers 11. fault tolerance This paper documents some of the design choices that have been made to address these challenges with emphasis placed on items 8 and 9. Much work has been done in the past twenty years to address many of the specific technical challenges listed in the previous section. Several references [2]-[7] provide excellent summaries of the advancements made by other teams competing in the 2005 DARPA Grand Challenge. The authors’ work related to the 2005 event is published in two references [8]-[9]. Numerous additional references can be cited for each of the important technical challenges. 2. SYSTEM ARCHITECTURE A hybrid Toyota Highlander was selected as the base platform for the system. Steering, throttle, braking, and transmission controls Figure 1. NaviGATOR