In regions with lots of snow, keeping roads clear is obviously a high priority. Luckily, the challenge involves infrastructure technology that’s undergoing a great deal of innovation. However, integrating that technology effectively and safely while making the best use of public funds is a tricky balancing act.
“Transportation departments face increasing pressure to keep roads open in all conditions,” says Mallory Crow, a researcher and graduate assistant at University of Akron in Ohio. “But the public doesn’t want to pay higher taxes. Economic cases can be made for multiple plows, tow plows, automated salt hoppers and hydraulics, automatic vehicle location and tracking, GPS, etc. Public agencies can’t invest in every technology, so they must make choices.”
In her doctoral thesis, Data Analytics for Winter Maintenance Decisions, Crow uses available data and research to help managers make the best possible choice as they weigh:
- Material costs. “It’s hard to predict the market and forecast seasonal needs. There’s also the environmental aspect of salting operations. In response, states have been investing in automated spreading, like Epoke North America’s EpoSat system, which uses GPS, software, and automated hardware to optimize spreading based on route location. This can save salt and increase public safety, but being sure it’s the right investment requires careful analysis.”
- Safety concerns. More technology means more equipment to monitor, which adds up to more stress for plow operators who, by definition, already work in difficult conditions. “It’s a lot to ask,” says Crow. “Some states say it’s hard to get drivers to take on the extra responsibilities.”
That’s the state of winter operations today: Technology that can improve service must be deployed cost-effectively without endangering employees. A system developed at the University of Minnesota as part of the Federal Highway Administration’s Intelligent Vehicle Initiative has been tested in Alaska, California, and the Midwest. Depending on your climate and weather patterns, it might work for your agency.
Apply automotive technology to plows
A driver assistance system combines several technologies into one system that monitors vehicle location and warns operators when they move off course or approach obstacles obscured by snow.
- Differential, real-time kinematic (RTK) GPS continuously tracks location. GPS receivers in the vehicle access RTK networks that are common across most of the U.S. as well as on-route reference stations. Precise mapping of lanes and other road elements is required.
- Collision avoidance technology based on light detection and ranging and/or radar sensors mounted on the vehicle, usually the roof.
- A cab-mounted head-up display synthesizes GPS and radar data streams to show oncoming vehicles and other potential obstacles and continuously compare vehicle location to lane and shoulder markings. Visual signals, such as changing centerline colors, warn operators they’re drifting off course or approaching an obstacle. Ideally, the system would ensure safe and effective plowing in zero visibility so operators can work during storms.
- Audible and tangible (haptic) alerts. Beeps and buzzers sound when obstacles loom; seats are rigged to vibrate when the vehicle approaches lane and road edges.
None are especially advanced in themselves. Most smartphones have GPS and Tesla electric cars come with radar, after all. But combining them into a workable system that improves safety and effectiveness in previously unthinkable conditions is.
Driver-assisted plowing was developed starting in 2001 by the University of Minnesota’s Intelligent Transportation Systems Institute (since 2013, the Roadway Safety Institute), a national University Transportation Center funded through federal transportation legislation. Intended for specialty vehicles like ambulances and patrol cars, the system was tested by the Alaska Department of Transportation and Public Facilities in 2003 and in 2012 by the California DOT (Caltrans).
Alaska: unqualified success
The institute helped Alaska employees outfit a blower and a plow to see if the system could help operators clear a 2,800-foot gap in the mountains northeast of Valdez where high snowfall rates and wind routinely create whiteouts.
Thompson Pass is the snowiest spot in Alaska, averaging 46 feet annually. In the winter of 1952–53, the most snow ever recorded in one season at one location in the U.S. fell: 81 feet. No attempts were made to keep the pass open during winter before then; today, because the road serves a port and a Trans-Alaska Pipeline endpoint, it must stay open all year.
Plowing the pass used to be costly, dangerous, and somewhat crude. Operators stayed on the road by scraping trucks against guardrails, which then had to be replaced. The agency tried a system that gave operators a heads-up when their truck traveled over magnets placed in the road, but not having a visual reference caused them to overcorrect their position.
After driver-assistance installation, plowing performance improved with less damage to guardrails. Quoted in a 2012 Government Technology article, “Smart Snowplows Keep the Highway to Valdez, Alaska, Clear”, Maintenance and Operations Chief Mike Coffey said, “It allows us to stay ahead of things rather than having the guys speed up when the storm’s over. If you have a need that justifies the expense, it’s a fabulous technology.”
Design, fabrication, installation, and operator training cost $136,000. In 2011, the agency spent an additional $553,000 to equip two more plows and a blower, upgrade the pilot project’s two trucks, and install a GPS base station. Today, one truck clears Valdez and half the road to the pass and the other trucks handle the rest of the road and the pass.
California: qualified success
“We heard about Alaska’s success and that led to a research project on Interstate 80, our main snow corridor here in California,” says Caltrans Supervising Transportation Engineer Joe Horton, PE.
The system also was tested on the 9,943-foot Tioga Pass in the Sierra Nevada mountains, one of eight passes typically closed in winter. The University of Minnesota collaborated with MTS Systems Corp., a sensor and measurement instrumentation developer in Evanston, Ill., to outfit two Caltrans plows and one blower.
Driver-assistance technology doesn’t recognize road components unless they’ve been digitized. Therefore, extensive mapping to support GPS location was conducted first.
Using a global navigation satellite system (GNSS) base station at a Caltrans garage, MTS employees drove along and mapped a section of Interstate-80 with an RTK GNSS receiver. This produced a highly accurate digital map with lane markings, pavement edges, guardrails, large signs, and median crossings for plow turnaround. Drainage inlets were added later for blower operations using point cloud data from laser scanning conducted by Caltrans.
The plows and blower were equipped with electronically scanning radar (ESR), an off-the-shelf machine-vision product from Delphi Automotive PLC of Washington, D.C. The sensor simultaneously scans for objects in two places: within 197 feet of the vehicle and at longer range, 656 feet from the vehicle.
In many ways, the system was a success. Operators found the head-up display excellent and useful. “It definitely has merit,” Horton says. “Also, the seat shaking is a really great feature that our drivers liked.”
But a serious weakness was also revealed.
Occasional ice buildup on the front of the roof-mounted radar enclosure produced a thin layer of water that absorbed most of the sensor’s radio signal, rendering the radar ineffective. Because neither Alaska nor Minnesota experienced a similar issue, Caltrans concluded the problem was caused by the type of snow prevalent in the Sierra Nevada.
“It’s a shame; the radar was a part of the system that was attractive to us and would have been very useful,” says Horton. “But that’s the nature of pilot projects. Driver assistance is still in the teething stage, and it turns out the wetter snow here compromised the radar. We’re hoping for a fix as the technology continues to develop, but for now it’s not reliable enough for us.”
The problem provides an irresistible metaphor: When developing new technology, it can be hard to see all the obstacles ahead. For now, driver assistance has proved itself in some, but not all, conditions. Still, it’s a promising technology and one to watch closely.