Page 26 - Penn State Civil and Environmental Engineering: Annual Report

LARSON TRANSPORTATION INSTITUTE NEWS LARSON TRANSPORTATION INSTITUTE NEWS Larson Transportation Institute
Optimal decision making for deteriorating roadway system
A team of of researchers from Penn State George Mason University University and and the University University of of Delaware explored how advances in in in in in ubiquitous sensing and and artificial intelligence (AI) could be used to to enhance decision making regarding the nation’s aging roadway system
The team created a a a a a a a a a a a a model for assessing when to inspect maintain and and repair roadway pavements and and bridge decks Even limited maintenance decisions have system-wide traffic impacts and are made with restricted budgets in in in in in mind according to the the the researchers Managers need both information about the the the the state of of the the the the roadway system
system
and a a a a a a a a a a a a a a system
system
for for decision-making that optimally weighs the the the options of of taking
or postponing actions including costs and construction downtime while also considering the benefits of an an improved roadway system
To this end the researchers collected crowdsourced vehicle response data data a a a a a a a a a a a potentially low-cost and high-volume high-volume data data source to detect roadway conditions especially in in in in high-volume high-volume traffic areas AI then identified deviations in in in in the the data indicating roadway defects With this information the the team created their decision-making framework using deep reinforcement learning a a a a a a a a a a form
form
of AI where an autonomous “agent” discovers the optimal sequential path of actions over over time The model accounted for randomness in in deterioration processes while also capturing the benefits of of improvements or or costs of of action postponements and and how they relate to to traffic impacts and and the the the larger roadway system
allowing the the the researchers to to consider many aspects of this complex problem “The developed techniques aid in in in in making cost-effective investment decisions over integrated roadway asset classes while explicitly recognizing traffic impacts to the the public and scheduling with them in in in in mind ” said Elise Miller-Hooks professor and Bill & Eleanor Hazel Chair in in in Infrastructure Engineering at George Mason University A full list of contributors to to this project can be found in the report Annual conferences return to in-person delivery
The fourth annual Transportation Asset and Infrastructure Management
(TAIM) Conference returned to to an an an an an in-person delivery
in in 2022 The event was held on on on October 17-18 at at the Wyndham Garden State College Designed for federal state and and local transportation agency representatives contractors and and and material suppliers engineering consulting
firm professionals faculty and and students TAIM features a a a a a a a a a a variety of of topics The twenty-seventh annual Transportation Engineering and Safety Conference (TESC) was held virtually on on on December 8-10 2021 More than 600 participants attended four workshops and forty-eight technical sessions TESC is the mid-Atlantic’s authoritative source of of information on on on on pressing issues from some of of of the foremost experts in in in in in transportation today Continuing education units are offered to to conference participants Registration is is now open for The twenty-eighth annual Transportation Engineering and Safety Conference (TESC) TESC TESC is is the the Mid-Atlantic’s authoritative source of of information on on pressing issues from some of of the the foremost experts in in in transportation today More than 600 participants attended virtually last year For the the most efficient and and secure registration all attendees exhibitors and and sponsors are encouraged to to use the the online registration portal Registration deadlines vary More information including the the schedule is is is available on on on on the the TESC website 26 CEE NEWSLETTER • VOLUME 38 2022