Two summers ago, the Minnesota Department of Transportation installed electronic message boards on parts of Interstates 35W and 94 to help warn drivers of crashes and to recommend speed levels during periods of high congestion.

Now, MnDOT would like to use the devices — officially known as Intelligent Lane Control Signs (ILCS) — to advise drivers of sudden stopping or slowing of traffic.  Many crashes occur when drivers cannot react quickly enough to these changes.

The Minnesota Traffic Observatory (shown in the feature photo above) is developing a warning system to detect such problematic traffic patterns and issue automatic advisories to drivers.

Shock waves on I-94

A section of I-94 in downtown Minneapolis, where southbound I-35W and westbound I-94 converge, may have the highest crash rate in the state.

As shown in the video above, vehicles constantly slow down and speed up here during rush hour, which causes a ripple effect called “shock waves.”

“There’s a crash every two days,” said University of Minnesota researcher John Hourdos, whose students watched over a year’s worth of video footage to document every accident and near accident. “They’re not severe crashes — no one has died for as long as I can remember, and most happen at slow speeds — but they cause a lot of delays for the traveling public.”

When statistics were still being kept, this section of I-94 had the highest number of accidents in the state, with approximately 150 crashes and 400 near crashes observed in 2003.

Researchers developed a program 10 years ago to detect “shock wave” patterns in the traffic, but they couldn’t develop a practical solution until the state invested in electronic message boards.

The University of Minnesota deployed cameras and sensors on three downtown rooftops in 2002 to observe traffic patterns.  They provide seamless coverage of the entire area, allowing researchers to watch vehicles from the moment they enter and exit the area. MnDOT has added additional cameras and detectors to watch over this roadway section. For the past year, the combined efforts of MnDOT and the university have provided data from 26 cameras and 12 traffic sensors for the two-mile section that includes the high-crash frequency location.

Thanks to the message boards, Hourdos and his team can now create an automated system to warn drivers when conditions for “shock waves” are greatest, using an algorithm he developed in the previous study.

Crosstown interchange 

A newer problem that researchers hope to tackle is the lineup of cars on I-35W southbound during rush hour at the newly reconstructed Crosstown interchange.

Although two lanes of traffic are provided for eastbound Highway 62 at the I-35W/62 split, these vehicles must later converge into one lane, due to the Portland Avenue exit. This causes a back-up on the 62 ramp that stretches back to 35W.

Hourdos said developing an algorithm to detect these queues is a different problem than what goes on with I-94, since there is a constant stoppage of cars and no rolling shockwaves.

“Combining the two methodologies will form a more robust solution and a single implementable driver warning system,” Hourdos said.

Researchers might target other problems areas should the state  install additional ILCS message boards elsewhere in the Twin Cities.