Dynamic Message Signs Displaying Weather Alerts Based on Roadside Pavement Sensors Were Associated with Up to a 1.5-mph Reduction in Average Speed and 2.0-mph Reduction in 85th-Percentile Speed.

Field Study Included Statistical Analyses on Traffic Data Collected from Sensors Upstream and Downstream of Dynamic Message Signs on US Route 12 in Minnesota.

Date Posted

Evaluation of Road Weather Messages on DMS Based on Roadside Pavement Sensors

Summary Information

Dynamic Message Signs (DMS) are often used to inform drivers about real-time road conditions and allow them to adjust their driving behavior based on the information displayed. Researchers analyzed traffic behavior under various winter weather conditions when advisory messages were displayed on US Route 12 between Delano and Maple Plain, Minnesota. A DMS was deployed at each upstream end of the corridor displaying messages triggered by roadside pavement sensors, in conjunction with eight warning flashers (Figure 1).

Photos of Slippery Road Sign with Flashing Light and Dynamic Message Sign Message: Slippery Pavement Use Caution

Figure 1. Advisory messaging and warning flashers (Source: MnDOT)


Data related to 14 winter weather events, gathered during the 2020-2021 winter season, were used to analyze driver behavior. To evaluate the effectiveness of the system on influencing driving behavior, multiple performance measures were established such as:

  • change in mean speeds,
  • change in 85th percentile speeds,
  • standard deviation in speeds, and
  • following distance / gap between consecutive vehicles

Data from temporary traffic sensors installed upstream and downstream of the DMS were used to analyze measures during winter weather conditions and baseline conditions. The upstream sensors served as the control location, reflecting no DMS messaging, while the downstream sensors were placed at a location potentially influenced by DMS advisory messaging. Speed data were aggregated every five minutes to calculate the mean speed, 85th percentile speed, and standard deviation. Individual vehicle-following behavior data such as headway was collected and then aggregated every five minutes to record the average gap, percentage of vehicles with a gap less than 1.5 seconds, and percentage of gaps less than 10 seconds. Winter weather events were identified based on the activation of the DMS messaging. Statistical analyses were used to assess the distribution of measurements and significance of effects.


  • In the westbound direction, mixed results were observed for the mean and 85th percentile changes in speed. Researchers suggested this could be attributable to external factors, as the westbound direction had the DMS located within the city of Maple Plain. Additional external factors included the presence of multiple intersections between the DMS and the downstream sensor, a change in maintenance district boundaries and road conditions between the sensors, and other factors which the study could not control.
  • For the eastbound direction, results from the evaluation of individual weather events indicated that 12 of the 14 events had statistically significant decreases in speed, averaging 3.5 mph in the subset. Additionally, a statistically significant decrease in the 85th percentile speed was found in 13 of the 14 events, averaging 2.9 mph in the subset.
  • Combined event analysis for the eastbound direction showed that the results for mean speed and 85th percentile speed were consistent with the individual events evaluation, with reduced speed downstream when accounting for the control period. The combined analysis showed the mean speed downstream was reduced by 1.5 mph and the 85th percentile speed was reduced by 2.0 mph.
Goal Areas
Results Type
Deployment Locations