Last March, one of the strangest races was held in the Mojave desert between Los Angeles and Las Vegas. Fifteen robot-controlled vehicles tried to make it almost 150 miles across the desert, and none of them did. The best of them traveled slightly more than seven miles.
Is there a control engineer anywhere who doesn't think he or she could have done a better job? You folks have dealt with Autonomously Guided Vehicles (AGVs ) for 20 years, designed control loops that can operate when networks and distributed control systems go down, and are holding your aging process plants together with spit and chewing gum as maintenance money disappears overseas. You can do anything. Of course a control engineer can do a better job. And maybe we'll just prove it to the world next year. Already, I have the Stealth Boys thinking about it. More on that later.
Robot racing came about because DARPA (Defense Advanced Research Projects Agency) has a mandate from Congress to develop unmanned vehicles for the military. Back in 2001, Congress said it wanted "the armed forces to achieve the fielding of unmanned, remotely controlled technology such that by 2015, one-third of the operational ground combat vehicles of the armed forces are unmanned."
DARPA has gotten nowhere with its usual defense contractors. After pouring a rumored $40 million into projects to develop a robot-controlled vehicle, DARPA decided to turn it over to the civilians. It joined forces with SCORE International, a Southern California based offroad racing organization, set up a 200-mile course across the Mojave desert, and announced the DARPA Grand Challenge (www.darpa.mil/grandchallenge ) to the robot world.
Is there a control engineer anywhere who doesn't think he or she could have done a better job?
The Grand Challenge offers a prize of $1 million to the first autonomous ground vehicle to complete a 142 mile route across the desert within a 10-hour time limit. The vehicle must navigate the course by itself without any outside assistance or remote control.
Over the past 12 months, challengers built about two-dozen very weird cars, ranging from motorcycles to six-wheelers. All had to undergo several days of qualification testing at California Speedway in Fontana. DARPA culled the list down to 15 qualified contestants, some of which were conventional 4x4 pickups, while others were home-built vehicles.
The actual course was unknown until three hours before the start, when contestants were given a CD containing the latitude and longitude of approximately 2,000 waypoints and speed limits along the route. The vehicles were supposed to find their way across the desert by using ground positioning systems (GPS) and vision sensors.
Alas, none of them were able to finish the course. Two failed to start, three completed less than a mile, and others suffered mechanical maladies. A few on the other hand, did quite well. Carnegie-Mellon University's Red Team made it the farthest, 7.4 miles, until its Humvee-based entry went astray, by getting caught on a berm in switchbacks on a mountainous section of the course.
The SciAutronics II entry from Thousand Oaks, Calif., a dune buggy built by TomCar Engineering in Israel and incorporating what appeared to be millions of dollars worth of support, made 6.7 miles before it slid off an embankment and got stuck. The SciAutronics I went off course at mile 0.75.
As you might expect, participants came mostly from robot companies, aerospace contractors, and universities. However, a few sponsors joined from the control world: Rockwell Automation, Green Hills Software, Measurement Systems, Intel, Rockwell Scientific, and SICK Sensors.
What the race really needs is a group of hungry engineers and mechanical wizards who understand controls and racing. Fortunately, I belong to such a group: The Stealth Boys.
The Stealth Boys all own Mitsubishi 3000 GTs or its Dodge-marketed equivalent, the Stealth, and we help each other out on various projects. The DARPA Grand Challenge might be right up our alley. Our group has a real time programmer, several electrical engineers from Rockwell Collins, high school kids, welders, electricians and fabricators"an eclectic mix who spend their time building street racers and road race cars.
So, if not us, why not one of you? To win the DARPA Grand Challenge, you need to keep it simple, keep it reliable, make it fast, and make the little bit of money you have for the project go as far as possible. You have been doing this for the past several years in your own process plants. Why not enter the Challenge and show all of us and the world what you've learned?