October 04, 2017, BY Matt Hughes

Autonomous vehicles are a frontier in transportation technology that now seems closer than ever, thanks to massive investments by companies like Tesla and Google as well as regulatory and transportation-planning changes by the federal and state governments.

But there are still fundamental challenges that must be overcome before fully autonomous, or self-driving, vehicles can traverse America's highways and city streets, says Bucknell University Professor Craig Beal, mechanical engineering, who specializes in automotive safety systems.

Now, thanks to a grant from the National Science Foundation, Beal is poised to tackle one of those questions, with research that can both improve existing vehicle safety systems and inform the development of self-driving cars.

The prototype electric vehicle Beal uses in his research allows computerized control for all major driving systems — except for the brakes. 

The $154,000 Major Research Instrumentation Grant awarded to Beal in September will provide equipment for Beal to measure steering torque — the effort required to hold the front wheels at a given angle as a vehicle travels around a turn — in hope of using those measurements to predict the friction coefficient, or grip, between the tires and the roadway.

"One of the questions that we want to answer is what exactly is going on in the steering system," Beal said. "If we understand that really well, then based on those measurements we should be able to estimate the friction coefficient of the surface we're on, so the car will know if it's on a wet or icy road, and how much grip it has."

He compared it to developing a computer program that senses what racecar drivers can feel behind the wheel — they can tell that they are pushing the car close to the limits of their tires' grip, Beal said, when the steering becomes lighter.

Beal drives the prototype electric car he will use in his research across the Bucknell campus. His actual experimentation will take place on a closed test course.

The results of his research could be used to improve the traction and stability control systems already present on most modern cars and inform the intelligent highway safety systems now in development alongside autonomous vehicles. For this reason, Beal believes both automakers and federal and state transportation agencies will have a strong interest in his results. When a more affordable version of the monitoring system he's using eventually makes its way into production vehicles, he said, the next questions for carmakers and transportation planners will involve how to network and share that information to make all drivers safer.

"If you encounter an icy spot on the road, that's good information for you to have, and you can react to it," Beal explained. "But it's even better information for the car behind you to have — it can slow down before it even gets to the icy spot."

Beal plans to make his measurements using sensors within the wheels of a prototype electric car that was built by students at Stanford University, where he earned his doctorate. Now housed in Bucknell's Dana Engineering Building, the car's steering and throttle are both controlled by an onboard computer, although the brakes have been intentionally left manual. The grant will fund the purchase of the in-wheel sensors, which cost about $70,000 each. (He'll need two, one for each front wheel.)

The sensors Beal will employ in his research will sit inside the wheel wells of his electric vehicle. 

At a closed test course at nearby Penn State University, Beal will drive the vehicle in specific maneuvers to measure steering torque through the full range of vehicle operation. He plans to involve student researchers in all stages of his research process, from helping design mounts for the sensors and electronics on the vehicle to interpreting collected data and even coming along for a ride.

"I've had students participate in testing as passengers," he said. "I can't let them drive the car out of safety and regulatory concerns, but they can participate as passengers and observers."

He plans to begin testing in the spring, and to continue working with the data he collects for several years.