In the foreword, Tony Tether, a past director of the Defense Advanced Research Projects Agency (DARPA), notes that DARPA’s role is to show that what is thought to be impossible can be done. In the field of robotics, the main rationale behind DARPA’s support of autonomous vehicle technology research is a government mandate, issued in 2000 [1], which aims to convert one-third of military land vehicles to unmanned units by 2015.

To achieve the required proof of concept, the agency organized the DARPA Grand Challenge series of robotic races. The first two challenges dealt with driving an unmanned vehicle in unknown rural and deserted environments. In 2004, 15 participants produced no winner, but in 2005, the race drew 43 participants, including Stanford’s winning Stanley vehicle.

This book describes the 2007 Grand Challenge--also known as the Urban Challenge--which proved to be the most complex yet. The vehicles were required to operate in urban environments (the housing area at George Air Force Base, in California), which included both autonomous vehicles and cars driven by humans. To make things more difficult, the autonomous vehicles had to show that they could pass California’s driving test, maintaining appropriate behavior and avoiding traffic infractions.

Editors Christopher Rouff and Mike Hinchey present the technical approaches and lessons learned by the 11 teams that qualified for the final event of the Urban Challenge. The book consists of five parts. Part 1 is an introduction and Parts 2 through 5 cover system and software architectures, navigation, control and sensors, and development and testing. Specific topics include planning, perception, decision making, control, interprocess communications, process management (including real time), data logging, and data fusion, as well as algorithms for obstacle avoidance.

The main contribution of the book rests in its descriptions of the different teams’ architectural options and technical approaches--some more traditional, some more innovative. For the record, the team from Carnegie Mellon University won the 2007 challenge; the previous challenge’s winner, Stanford, came in second. The Virginia Tech team’s car also completed all of the missions; it placed third.

This book is not overly technical and will be of use to anyone interested in learning more about the rapidly growing field of robotics--don’t tell me your vacuum cleaner is still hand-driven! We should be looking forward to the next challenge, this time focused on humanoid robotics, which is set to launch in October 2012 and run for approximately 27 months [2].