Autonomous vehicle challenge led tnew technologies and invigorated the prize challenge model of promoting innovation
At the break of dawn on March 13, 2004, 15 vehicles left a starting gate in the desert outside of Barstow, Calif., tmake history in the DARPA Grand Challenge, a first-of-its-kind race tfoster the development of self-driving ground vehicles.
The immediate goal: autonomously navigate a 142-mile course that ran across the desert tPrimm, Nev.
The longer-term aim was taccelerate development of the technological foundations for autonomous vehicles that could ultimately substitute for men and women in hazardous military operations, such as supply convoys.
The Grand Challenge was designed treach beyond the traditional defense performer base and tap intthe ingenuity of the wider research community. It was DARPAís first major attempt tuse a prize-based competition tattract novel performers and ideas and encourage collaboration across diverse fields. The first team tpass a series of qualification tests and then complete the course in less than the prescribed ten-hour time limit would receive a $1 million cash prize.
The technological hurdles and rugged desert course proved tbe tomuch for the teamsí first attempt. None finished the courseóthe top-scoring vehicle traveled only 7.5 miles and the prize went unclaimed. The competition wasnít a loss however; it offered a promising glimpse at what was possible.
That first competition created a community of innovators, engineers, students, programmers, off-road racers, backyard mechanics, inventors and dreamers whcame together tmake history by trying tsolve a tough technical problem,said Lt. Col. Scott Wadle, DARPAís liaison tthe U.S. Marine Corps. ìThe fresh thinking they brought was the spark that has triggered major advances in the development of autonomous robotic ground vehicle technology in the years since.
Just one day after the first challenge ended, DARPA announced it would hold a second Grand Challenge in the fall of 2005, 18 months after the first. This time, after analyzing lessons learned, five vehicles out of the 195 teams that entered successfully completed a 132-mile course in southern Nevada. Stanford Universityís entry, ìStanley,finished first with a time of 6 hours and 53 minutes and won the $2 million prize.
Tfurther raise the bar, DARPA conducted a third competition, the Urban Challenge, in 2007 that featured driverless vehicles navigating a complex course in a staged city environment in Victorville, Calif., negotiating other moving traffic and obstacles while obeying traffic regulations. Six teams out of 11 successfully completed the course. The ìTartan Racingteam, led by Carnegie Mellon University, placed first in points awarded based on time tcomplete and ability tfollow California driving rules and won the $2 million prize.
Although it isnít easy tquantify the effects of these DARPA challenges on the development and deployment of autonomous vehicle technology, 10 years later defense and commercial applications are proliferating. The rapid evolution of the technology and rules for how tdeploy it are being driven by the information technology and automotive industries, academic and research institutions, the Defense Department and its contractors, and federal and state transportation agencies. Within DOD, some of the efforts timprove upon and deploy autonomous ground vehicle technology include:
- Oshkosh Defense developed the TerraMax unmanned ground vehicle for the Marine Corps;
- TORC Robotics, one of six finishers of the Urban Challenge, continues tdevelop utility-vehicle-scale autonomous capabilities for Marine Corps platforms;
- Today, three other DARPA challenges are building on the DARPA Grand Challenge prize-based competition model:
- The Spectrum Challenge is a competition tdemonstrate a radiprotocol that can best use a given communication channel in the presence of other dynamic users and interfering signals, with a goal of enabling reliable communication in a congested environment. The challenge includes head-to-head competitions between competing teams in a structured test bed environment. The Spectrum Challenge Finals will occur March 19-20, 2014 at DARPA.
- The DARPA Robotics Challenge has a goal of developing robots for use in responding tnatural and man-made disasters. The DRC was structured with three planned competitions tallow teams time tincorporate lessons learned from one event tthe next. Teams are currently preparing for the third and final planned competition, the DRC Finals. Strong performance by the teams in the second event, the December 2013 DRC Trials, encouraged DARPA tincrease the difficulty of the final round. The results of the DRC Finals will give DARPA, DOD, and industry a preview of what is possible with robots.
- The Cyber Grand Challenge is a tournament for fully automated network defense. Similar tcomputer security competitions played by expert software analysts, the CGC will require automatic systems treason about software flaws, formulate patches, and deploy them on a network in real time. The CGC aims tunite program analysis experts with the computer security competition community tbring automation research out of the lab and intthe field.
DARPA expects that, like the original Grand Challenge before them, these challenges will encourage new waves of research and development that will spur continued innovation, encourage commercial investment, and lower the cost of advanced technologies.