Putting Racecar Data Into Consumer CarsBy Baselinemag | Posted 2002-11-01 Email Print
How Real-World Numbers Make the Case for SSDs in the Data Center
Honda Racing takes data from its racecar drivers and puts the results into its consumer vehicles.
Brazilian Formula One driver Tony Kanaan sticks his foot onto the throttle of the 800-horsepower Honda engine behind him, as he pulls into Turn 11 of the Long Beach Grand Prix.
G-forces push him back in his seat. Imperfections in the asphalt vibrate the steering column. He plunges into the long straightaway, flying past the start-finish line at 190 mph. Colors at the edge of his vision begin to blur. "It's very fast out there," he says. "And it feels very lonely."
But he's being watched closely. In a pair of semi-trailer trucks crammed with millions of dollars worth of computer equipment, two Honda engineers near the pits watch the telemetric data flow from his car at the speed of light.
Suspension movements, spring tension, acceleration and deceleration, aerodynamic analysis of the downward thrust exerted by inverted "wings," heat buildup in brakes, heat-stress of metal components in wheel bearings, "steering trace," traction control, even air and fuel mixtures are all recorded into five massive databases.
During the two-hour race, a gigabyte of data comes from Kanaan's carand from each car in Honda's racing fleet. Every 30 minutes, 20 engineers in the semi-trailers dump data over a virtual private network to Honda R&D facilities in Los Angeles and Tokyo. They store information, collaborate and communicate via Lotus Notes. A Sybase system, custom-designed for Honda Racing, updates databases on the fly.
Fiber connections carry text messages, photos and raw encrypted data simultaneously, helping engineering teams and designerswho specialize in assemblies, braking systems and fuel monitorsimprove designs and spot weaknesses affecting production models. The team's five databases catalog engine management files, record time information and analyze statistics.
Team manager Tim Cunningham says the approach helps find solutions to complex problemslike overheating in a set of critical engine bearings.
"Engine bearings run at much higher rotational speeds," he says, "because engines are running much harder and faster than in production cars."
When that problem occurred, engineers had only days to fix it before qualifications for the next race. Such deadline pressure, experienced in two-year stints with racing, helps them solve problems faster when they return to production engineering.
The end-game for these mounds of data is far from the trackon a Honda dealer's showroom floor. Better traction that drivers feel but never see, better anti-lock brakes and other safety featureswithout significant price increasesare payoffs earned when customers drive a new Accord off the lot.
Honda uses the track as a test lab to apply methodical engineering to create roadworthy consumer automobiles. "We do know that for Honda, the race track is the proving ground for new engineers," said Kevin Prouty, automotive specialist for AMR Research.
But it's how the data gets used that matters. "It takes a good engineer to make real-world connections," he says. "For Honda, it is not the esoteric and experimental that is important, it's the translation to a commercially viable product."
The data from the track is not directly translated into the design of new cars, Honda American Motors officials note. Nor does Honda use as much data from end to end as do North American manufacturers such as GM, which layers on technology to drive down time from design-freeze to production.
Nonetheless, a highly engineered production process is what speeds the design-to-showroom times for models like the 2003 Honda Accord.