Chang, Students Honored with Mickle Award

news story image

From left: TRB Executive Committee Chair Nathaniel P. Ford, Sr.; Professor Gang-Len Cnang; Yi-Ting Lin; Yen-Lin Huang, and TRB Executive Director Victoria Sheehan, Chang, Lin, and Huang received a Certificate of Award at the TRB meeting’s Thomas B. Deen Distinguished Lecture and Presentation of Awards on January 9.

A team of University of Maryland (UMD) transportation researchers led by Professor Gang-Len Chang has won the D. Grant Mickle award, given each year by the Transportation Research Board National Academy for the best paper in the area of operation, safety, and maintenance of transportation facilities.

Corresponding author Yen-Lin Huang, Yi-Ting Lin, and Chang officially received the award for their paper, “Extending the I-95 Rule-based Incident Duration System with an Automated Knowledge Transferability Model,” at the 2023 Transportation Research Board annual meeting in January. Their work won the TRB Freeway Operations Committee’s Best Paper award and was then selected for the Mickle award from among 90 papers nominated by TRB committees. 

In their paper, the UMD researchers presented a method that can help expand the scope of a model used in highway incident response without a large increase in cost and resources.

Accidents and other highway incidents can lead to long delays for commuters and increase congestion. To mitigate these effects, many state highway agencies utilize Traffic Incident Management (TIM) systems, which allow them to detect, respond to, and clear traffic incidents with greater efficiency. An effective TIM system can reduce the clearance duration of detected incidents, and minimize the resulting impacts on traffic delay and safety.

As part of its TIM system, the Maryland Department of Transportation has made successful use of a rule-based incident duration prediction model (IDPM) that covers Interstate highways I-95, I-495, and I-695. In brief, the model draws from incident records to make predictions which can then be used in response planning and resource allocation.

With the model having proved effective, MDOT is now planning to extend it to cover highway systems throughout the state. Doing so, however, requires obtaining sufficient incident data to properly calibrate the model’s parameters–and harvesting that data can be both time and labor-intensive.

The method proposed by Huang and his colleagues could help overcome that challenge. In their paper, the researchers put forward a knowledge transferability analysis (KTA) method, making use of an automated process that can assess, select, and transfer existing prediction rules in order to estimate incident durations on a new target highway. 

Tests of this approach with two different data sets yielded accuracy levels of 82% and 87%, respectively. According to the authors, these rates are comparable to the current IDPM’s performance, but require far fewer records for model calibration.

"Most real-world transportation problems face insufficient data,” Huang said. “Our study sheds light on the power of such precious and hard-earned data. That is its main contribution.”

“It was an honor for us to receive the award and we feel grateful for being recognized. It motivates us to continue high-quality work and help solve additional transportation issues, thus helping to bring about a better transportation environment,” he said.

Professor Chang, likewise, emphasized the importance of building on the team’s success. "We are certainly pleased with the award, but I also remind my students that we should be humble and committed to doing further high quality work to bridge the gap between state of the art and state of the practice,” he said.

 The D. Grant Mickle Award was established in 1976. It honors the fifth TRB executive director, who was later appointed a member of the Board's Executive Committee and became its 33rd chairman. 

Huang, Lin, and Chang received their Certificate of Award at the TRB meeting’s Thomas B. Deen Distinguished Lecture and Presentation of Awards on January 9. 

Published January 19, 2023