Carnegie Mellon University's Driving Research Center (DRC) is using advanced technologies to enhance driving safety. Activities at DRC involve the development and testing of transportation related devices using a sophisticated driving simulator. The simulator allows researchers to test various systems in a safe and controlled setting.

Carnegie Mellon has long been a leader in research for the Department of Transportation's Intelligent Vehicle Initiative. Carnegie Mellon conducted the first research effort of its kind in developing a "Perclos camera" sensor system designed to prevent accidents caused by drowsy drivers. This dashboard-mounted camera monitors eye closure and measures the level of drowsiness. To properly test this system, it was necessary to monitor a drowsy driver operating a vehicle. Under normal operating conditions, this would have been extremely dangerous. The necessity to test this device led Carnegie Mellon to form a partnership with ISIM, the developers of TruckSim.

TruckSim is a unique research simulator providing a full-motion, hi-fidelity, virtual truck driving experience. What makes TruckSim a truly outstanding research and training tool is its ability to simulate a wide range of truck engines, transmissions, driving conditions and scenarios. It is a unique research laboratory that enables researchers to evaluate a driver's performance or determine their level of mental alertness in a safe environment. Data obtained from the Perclos camera and other sensors located in the simulator are processed with Carnegie Mellon's data reduction program.

The simulator consists of an actual Freightliner truck cab sitting atop a sophisticated motion platform which duplicates the pitching, surging, heave and rolling movements a driver would feel when operating a real truck. The front of the cab is surrounded by three wrap-around screens on which the driver's field of vision is projected. Smaller screens attached to the cab simulate the reflections in the rear view mirrors.

The virtual world of the driver can be altered by operators at a computer console. The operators can change road and weather conditions while the driver is at the wheel of the simulator. There are 30 different scenarios truckers typically face: day or night driving, changing traffic patterns, tire blowouts and brake failure, shifting or heavy loads, city streets and open highways, drunk or aggressive drivers, etc. TruckSim can simulate different environmental conditions - rain, snow, ice, and even factor in the behavior of the vehicle on slippery pavements.

The use of this leading edge technology creates an extremely accurate representation of what a driver experiences while operating a heavy vehicle. This has been confirmed by a number of truck drivers who were interviewed after completing a four hour simulation of their 12-hour shift on the simulator. This long work day resulted in a number of fatigued drivers and allowed Carnegie Mellon to obtain valuable sensor data without jeopardizing the drivers' safety.

TruckSim will be used in three major areas at Carnegie Mellon: sensor evaluation, human factors research, and curriculum development for driver training.

The first major research initiative builds upon Carnegie Mellon's past work in preventing accidents caused by driver fatigue. "Research that would be life-threatening under any other conditions can be done in complete safety with TruckSim," says John Tabacchi, Manager of the Transportation Research Center at Carnegie Mellon.

"Current research using TruckSim looks at the performance of tired or stressed drivers," he says. Drivers who have been driving for up to 12 hours are monitored to see how alert they are at the wheel. The data collected about driver reactions to various situations can be used to create new technologies aimed at improving road safety, such as alerts that will let a driver know when it's time to pull over for a rest stop.

As dashboard displays and driver information systems become more sophisticated and more widely used, drivers of the future will be bombarded with information. They'll have navigational displays, cell phones, test messages, sensors and crash warning systems to monitor. These safety tools can also present problems if they become too distracting. "The question is, what information does the driver need to operate the vehicle without getting information overload?" Tabacchi says. The simulator lets human factors engineers test new instrumentation and displays before actually installing them in working vehicles.

Simulators as training facilities aren't new. Airline pilots have trained on them for years. Similarly, truck simulators will enable beginners to hone their driving skills before they actually go out on the road. They can learn in a safe environment how to do the right thing in the event of a tire blowout or brake failure. In the real world, one would have only one chance and a few seconds to act: in the simulator, the trainee can afford the time to learn how to avoid dangerous situations.

The DRC is continuing to work with companies to improve driver safety. Safe Drive Technologies is developing a driver's education curriculum incorporating simulator technology. The goal is to use this technology to update and improve current driver training methods and the driving performance of young drivers. "Driving simulators can supplement current driver education programs and teach inexperienced drivers defensive driving techniques that can reduce accidents," Tabacchi says.

Conventional driver training methods teach how to drive under normal driving conditions. Actual experience in driving in hazardous conditions such as snow and ice, or with a flat tire, isn't possible because of the danger to student and instructor. With simulators, however, the drivers can be placed in these kinds of situations and taught how to deal with them correctly and safely.

Research indicates that drivers, especially teenagers, are prone to "the Fatal Five" driver errors: failing to yield, not adjusting to bad weather or darkness, tailgating, losing control, and driving under the influence. With simulators, they can see how these potentially hazardous situations can be avoided. For example, scenarios can be developed where students can experience other drivers failing to yield, pulling out in front of them or the dangers of tailgating. They can also experience how a car will handle in bad weather conditions or when a tire is low or flat.

Advances in computer technology are paving the way for low-cost, high-fidelity automobile simulators that could make widespread use of simulators in driver education programs feasible. "There is a need to reevaluate current driver training programs to ensure that they are up-to-date and taking advantage of all the technology that is currently available," Tabacchi says.

For additional information or questions contact: Jeff Hibner 412.268.3428 email jh8k@andrew.cmu.edu