Driver-assist systems support snowplow operations
Snowplow operators have many tasks to do simultaneously while removing snow and spreading deicers. They often do this challenging work in low-visibility conditions, for long hours at a time. In an extensive line of studies, U of M researchers have developed technologies to help snowplow operators perform their jobs more safely and effectively.
The most recent research, a multiyear project funded by the Minnesota Department of Transportation (MnDOT), studied driver-assist systems for three applications: gang plowing, backup assist, and lane-boundary guidance.
In gang-plowing operations (also called tandem plowing), one plow must follow another at a specified position offset. Led by Professor Max Donath, mechanical engineering (ME) researchers evaluated the performance and accuracy of onboard units commercially available in 2015–16 that enabled vehicle-to-vehicle (V2V) communication. “We found that using only V2V communication between two onboard units didn’t give the plow operators sufficiently accurate information to maintain spacing between two vehicles,” Donath says. “This will likely change in the future as higher-accuracy GPS modules become available.”
As the team was completing these tests, Donath learned from MnDOT’s Metro District maintenance staff that they were interested in a driver-assist system that would inform plow operators of the edge line of a road boundary, particularly when visibility is limited. In an additional request, the maintenance staff wanted the plow operator to be able to detect a vehicle behind the truck during backup. “As a result, the scope of our study was modified to address these needs,” he says.
The research team, which included co-investigators Chen-Fu Liao and Nichole Morris, developed a new lane-guidance system based on a high-accuracy differential Global Navigation Satellite System. It was installed on a snowplow operating on MN-25 running between Belle Plaine and Green Isle, a flat area that has the worst winter visibility in the Twin Cities metro.
The system is a less complex and more affordable version of the one developed in previous work and deployed in Minnesota and Alaska. “This new system omits the custom-made head-up display, which was a costly component,” says Liao, ME senior research associate. “The cost per plow of the old system was about $60,000, but the new system totals just over $15,000. We believe that the lower cost could lead to wide-scale deployment.”
Tony Johnson of the MnDOT Shakopee Truck Station adds that plow drivers really like the new system. “We’ve actually asked for more [systems] for other trucks,” he says.
To provide backup assistance, the research team developed a rear-facing radar-based system and deployed it on a snowplow that operates on Highway 169 in the southwest metro. When an object is detected, the system provides an audio warning to the operator to look at the display from a rear-view camera. “The plow operators say the system supports their work in crossovers, the areas where they cross divided highways,” Liao reports. “Crossover operations often require backing up, which can cause safety concerns.”
Morris, director of the U’s HumanFIRST Lab, led studies of human factors issues to better understand operator needs prior to designing the systems. “For example, we learned that drivers didn’t want an additional graphic display for lane guidance, so our team made a very simple adjustable LED display to show relative position in the lane,” she says. Her team also observed how operators were using the system for safety and efficiency. “Excitingly, they were not only using the system to stay on the road, they were also using it to clear a straight-line path, eliminating the need for an additional passover to catch patches left behind.”
- Research project page
- Research report
- Video: CTS Research Makes an Impact: Driver-Assist Systems (3:11, Sept. 2018)