As Workload Overwhelms, Cars Are Set to Intervene

Share with others:


Print Email Read Later

Heading south on Route 34 toward Jersey Shore beaches on a summer weekend, drivers confront a daunting array of highway quirks, not limited to jughandle intersections and baffling exit signs.

The simple act of turning left on Allaire Road in Wall Township, for example, is confounded by a traffic circle, where an attempt to head east casts the driver into a ballet of choosing the proper lane, looking for the exit and maintaining a high alert in the crush of beach-seeking vehicles.

Now imagine that during this encounter a low-tire warning flashes on the dashboard. Next, a chime alerts the driver that a text message -- maybe important -- has landed. Then the cellphone rings.

The overload of inputs, perhaps amplified by foul weather or a demanding toddler, presents a real challenge to the driver -- and a danger to all road users. The National Highway Traffic Safety Administration estimates that distraction and inattention contribute to 20 to 30 percent of reported crashes.

Much as regulators and automakers have rushed to deal with the flood of distractions that invade the automobile -- GPS displays, Internet radio, e-mail and even Facebook apps -- there is a growing effort by engineers to build cars that gauge the difficulty of situations and recognize a driver in distress. Then the car would react, delaying all but the most urgent alerts, sending phone calls to voicemail and freeing the driver to focus on the task.

The study of driver workload management -- some would point to the irony in this reaction to a situation partly created by automakers themselves -- is progressing alongside the efforts of the planners who dream up new generations of infotainment features. A foundation of workload study is the Yerkes-Dodson Law, a theory developed in the early 20th century that plots workload and performance on a bell curve.

There can be trouble at either end -- an inattentive, underworked driver may be as much a risk as an overworked driver who cannot handle the combined sensory inputs and driving chores. In the middle is the ideal, a driver functioning at optimum level.

Systems that detect driver drowsiness, like the Mercedes-Benz Attention Assist feature, can prompt a driver to be more alert, but driver overload is harder to manage. N.H.T.S.A. has issued voluntary accessory-design guidelines in an effort to reduce distraction, but given consumers' hunger for gadgets, managing those distractions to reduce workload may prove a better solution.

As safety groups press for restrictions on phone conversations and messaging in the car, the urgency to find a solution will only increase, experts say. Studies of driver workload have a long history, but a milestone came in 2003 when the John A. Volpe National Transportation Systems Center, a unit of the Transportation Department -- in conjunction with Delphi, the giant parts supplier, Ford Motor and several universities -- began a research project to quantify distractions and driving situations as a way to generate workload estimates.

Paul A. Green, a University of Michigan research professor, said in a telephone interview that the Volpe study stimulated research. Today, automakers and universities are developing technologies that will let them measure the level of driver stress and the response to the pressure. That data could be used by a management system that would delay calls, alerts, text messages and warning lights at the times when the driver's workload was peaking and the stress level was high.

Because many cars are equipped with advanced electronics -- radar, sensors and cameras developed to enable features like smart cruise control and lane-departure warning -- some of the equipment needed to gauge workload is already in vehicles. Sensors that determine speed, throttle position, steering wheel angle and transmission gear selection, and even weather conditions, can be adapted to see traffic on the road and monitor driving situations.

Jeff Greenberg, a Ford senior technical leader, said in an interview that his research team had built models that predicted workload based on information from data points relative to time. The model takes into account that when steering and throttle position are constant for an extended time and outside sensors show moderate traffic, workload will be minimal. But if over just a few seconds the accelerator is pressed, traffic becomes heavy and the wheel is turned, the system determines that the driver has encountered a changing situation and workload is increasing.

Those elements describe the driving situation. But what about the driver?

That is where biometrics -- devices that measure how well the driver is managing workload -- come into play. A simulator built by Ford to demonstrate biometric tools uses sensors in the steering wheel, like those on exercise machines, to monitor heart rate. Sweaty palms are detected by gauging skin conductivity. Probes aimed at the driver's face measure skin temperature, and a sensor in the seat belt can tell when the driver is breathing hard.

Mr. Greenberg said that by combining biometric data with an analysis of the driving situation, workload could be very accurately gauged and distractions could be delayed until things were on an even keel.

Some automakers have not embraced biometrics as a practical way of measuring driver stress. Jim Foley, senior principal engineer at Toyota's research center in Ann Arbor, Mich., said in a phone interview that although work on biometrics started in the 1970s and continued, there were problems in applying the technology to production vehicles. He added that measuring anything through the steering wheel did not work if the driver was wearing gloves and that getting a reliable signal for all drivers was challenging.

Bryan Reimer, a scientist engaged in driver workload studies at the Massachusetts Institute of Technology, disagreed. "Some researchers believe biometrics work in the laboratory only," he said in a phone interview. "Others, including myself, believe firmly they can be used in the car."

He added that heart rate could be measured through the seat as well as through the steering wheel. "We are doing a considerable amount of work using biometrics in conjunction with other visual information, such as eye tracking and vehicle telemetry, to provide a holistic view of the demands the driver is under," he said.

Honda and Ohio State University are collaborating on a driver workload study. They have built a simulator on a platform that mimics vehicle motion while driving scenarios are projected on a wraparound screen and driver performance is charted. Biometric sensors measure blood pressure, heart rate, eye movement, respiration, blink rates, skin conductivity and, by means of functional near-infrared technology, brain activity.

"We're trying to get a better understanding of how the driver thinks about the system while driving the vehicle," said Steven Feit, Honda's chief engineer for infotainment research. "We don't want to just lock out features, but we want to make them available at the right time. We're trying to do a balanced approach between controlling the environment and optimizing the operation of infotainment features."

Janet Weisenberger, director of the Ohio State University simulation laboratory, said that this research could be applicable to the development of semiautonomous vehicles. "In the long term, you want to know things about the driver so systems could be best matched to a driver's capabilities."

The semiautonomous car that can assist the driver in difficult situations is an obvious next step beyond using workload detection to manage distraction. If the automobile can gauge workload, it can lend a hand when necessary.

Most luxury cars are available with collision avoidance systems and intelligent cruise control, and, as workload measurement technology is refined, other aids will become possible. For example, if sensors see that snow is obscuring the road and biometrics indicate driver stress, the vehicle could project lane lines on the windshield. But there might be limits to how much control automakers would want to wrest from the driver: leaving the driver with little to do could be problematic.

"Lost in the debate around autonomous technology," Mr. Reimer of M.I.T. said, "is that these features pull workload away from the driver and can result in under arousal. The traditional cause of that is driver fatigue, but semiautonomous features may be driving us in that direction."

How soon might we see a system that combines biometrics and telemetry to manage workload? Alan Hall, a Ford spokesman, said in e-mail, "We have not explicitly stated when you could see the workload detection system in a car, but as a project in the research phase, the earliest time frame would be three to five years."

Ford, of course, may not be first to market with workload management. "All automakers are working at different levels in this area," Mr. Reimer said. "And there's a knowledge that first to market wins."

autonews

This article originally appeared in The New York Times.


Advertisement
Advertisement
Advertisement

You have 2 remaining free articles this month

Try unlimited digital access

If you are an existing subscriber,
link your account for free access. Start here

You’ve reached the limit of free articles this month.

To continue unlimited reading

If you are an existing subscriber,
link your account for free access. Start here