Integrated Modeling for Road Condition Prediction — Phase 3 Evaluation Report
Chapter 5. Conclusion
This evaluation has resulted in a better understanding of the impact the Integrated Modeling for Road Condition Prediction (IMRCP) Phase 3 had on Kansas City Scout (KC Scout) operations, the accuracy of the speed predictions and forecasts, and the usefulness of the information to KC Scout operators and supervisors.
Accuracy and reliability of the IMRCP traffic speeds can affect the operator’s confidence in and the use of IMRCP’s traffic-predictive applications. The Traffic Estimation and Prediction System (TrEPS) predicted speeds were inaccurate on several of the observed dates. This model was found to be sensitive to missing or erroneous input parameters, such as changes to posted speed limits, roadway construction (or blockages), and erroneous or missing detector data. The machine learning-based prediction (MLP) speeds was often found to be accurate and within about 5 miles per hour (mph) of the detector speeds. MLP was less affected by the construction work zone changes at Interstate 435 (I–435) at State Line Road, but occasionally had periods with large deviations (up to 20 mph) above or below the detector speeds. Forecast speeds for MLP and TrEPS showed periods where the forecasts appeared to have large deviations (up to 20 mph) above and below the detector speeds. With MLP speed forecasts, the shorter-term forecasts tended to be more accurate than the longer-term forecasts.
Operators reported not using the traffic predictions capability, because they preferred to use existing tools for obtaining information about real-time traffic conditions and incidents. Based on the project experience, traffic management center staff were focused on current traffic events and conditions, monitoring known problem areas and incidents. They often did not have the time or means to address the overall network conditions and assess the IMRCP traffic predictions. There seemed to be a preference from the operators to have IMRCP outputs integrated into existing TMC interfaces.
Although IMRCP traffic speed predictions had minimal operational impact during the 2018–2019 and 2019–2020 winter driving seasons, operators reported that they often referred to IMRCP for weather forecast information prior to weather events. Interviews and discussions with KC Scout operators and supervisors revealed that operators liked the weather-related prediction components of IMRCP and primarily assessed it prior to a winter weather event. The operators said they liked the precipitation/weather predictions and the pavement status predictions, because they can provide some insight into what locations to monitor more closely as the weather moves in. The operators used this feature to find areas with potential issues affected by the incoming weather. They reported it helped them prioritize and focus attention on particular roadway sections or areas, which in turn may improve their communication and decisions relating to motorist assist deployments.
The findings from this evaluation may be helpful to inform future improvements to the IMRCP application and any new deployments. Additional information can be found in the Integrated Modeling for Road Condition Prediction Phase 3 Project Report which summarizes the lessons learned from the evaluation and the KC Scout demonstration, as well as potential applications for IMRCP and recommendations for further study.
United States Department of Transportation - Federal Highway Administration