University of Minnesota Driven to Discover
Universities MUniversities Wordmark
Center for Transportation Studies Heading

Webinar Available

Intelligent Decision Support Inside the Vehicle: Can it Help Drivers Make Safer Driving Decisions?

Caroline Hayes, Professor, Department of Mechanical Engineering, University of Minnesota–Twin Cities

September 9, 2010

A disproportionate number of accidents and fatalities occur at rural thru-stop intersections where a secondary road crosses a divided highway. Drivers trying to cross the highway or enter traffic are very vulnerable, especially when traffic is heavy. Intelligent decision-support systems may help drivers make better choices about when to enter the intersection.

At the September 9 Advanced Transportation Technologies seminar, Caroline Hayes, professor of mechanical engineering, presented the results of a study that compared three decision-support systems.

Two of the decision-support systems are based on visual interfaces. The first consists of an active LED icon-based sign in the intersection. A yellow rectangle indicates an approaching vehicle, and a red circle and red rectangle warn that the approaching vehicle poses a threat to drivers preparing to cross or merge. This system is currently being field-tested at the intersection of Highway 52 and County Road 9 in Goodhue County, Minnesota.

The second system consists of a modified LED icon display placed on the right- and left-side mirrors. The display features an approaching car with speed lines and a fill bar. The amount of color in the fill bar indicates the time to arrival of oncoming traffic.

These interfaces are easy to understand and provide drivers with a lot of information, but they rely on the visual channel, which is already overloaded because driving is a very visual task, Hayes said.

The third interface consists of a seat with two vibrating pads that provide vibrotactile feedback. A vibration in the right pad signals that traffic is approaching from the right; a vibration on the left indicates traffic approaching from that side.

Testing of the three systems was conducted in the HumanFIRST driving simulator with 24 subjects between the ages of 19 and 69. Subjects were evenly divided between male and female. Once trained, the subjects were asked to drive through a simulated thru-stop intersection under four conditions: 1) without using decision support (this was the control), 2) using the LED icon traffic sign, 3) using the LED side-mirror displays, and 4) using the vibrotactile seat.

The seat was more effective than either of the visual displays, Hayes said. Drivers were able to keep their eyes on the road and enter traffic with more space between their car and the oncoming traffic. But drivers didn’t like the vibrotactile interface because they found it “weird” or “annoying.” They also reported that they expected the interface to tell them when to “go” instead of when “not to go.”

Drivers preferred the icon sign in the intersection, probably because the location and the interface were familiar, Hayes said. The sign gave drivers nuanced information, but they also continued to look at it after a gap in traffic became available. This effectively reduced the safety margin.

Hayes pointed out, however, that although the vibrotactile seat provided a better safety margin than the icon sign, that margin was not significantly better than the safety margin when no decision-support system was used.

Going forward, researchers may wish to experiment with modifications of the vibrotactile seat, such as changing the vibration to make it less annoying or to provide more information. Increased training would probably make the seat more familiar and acceptable to drivers, Hayes said. But, she added, there are larger questions to address. For example, can intelligent decision-support systems really help drivers make better decisions? Right now, the answer to that question is “maybe.” Another question to consider is whether drivers really want their cars to tell them what to do. According to Hayes, “Some do, and some don’t.”