Strategic Prioritization and Planning for Multi-Asset Transportation Infrastructure Maintenance, Rehabilitation, and Improvements: Phase 2 - Data-driven Decisions from Continuous Monitoring

This project will extend prior work by this team in a first phase of this project, “Strategic Prioritization and Planning for Multi-Asset Transportation Infrastructure Maintenance, Rehabilitation, and Improvements: Phase 1 – Prioritization through Optimization.” This 2-year, first-phase project will wrap up in coming months. Outcomes of the first phase will include: •A deeper understanding of the nature of crowdsourced vehicle response data and its utility, specific to the perception of asset (roadway and bridge) condition, with its impact on free-flow speeds and capacities, and the ability to detect deteriorated conditions through latent space modelling of the data and developed machine learning algorithms; This project will extend prior work by this team in a first phase of this project, “Strategic Prioritization and Planning for Multi-Asset Transportation Infrastructure Maintenance, Rehabilitation, and Improvements: Phase 1 – Prioritization through Optimization.” This 2-year, first-phase project will wrap up in coming months. Outcomes of the first phase will include: •A deeper understanding of the nature of crowdsourced vehicle response data and its utility, specific to the perception of asset (roadway and bridge) condition, with its impact on free-flow speeds and capacities, and the ability to detect deteriorated conditions through latent space modelling of the data and developed machine learning algorithms; •Development of probabilistic predictive models for multi-asset (pavement and bridge) roadway system serviceability levels, with and without maintenance or other improvements, while considering inspection accuracy needs, activity impacts and other associated costs; •Conceptualization of the multi-asset, strategic planning of maintenance, repair and rehabilitation options (improvement actions) and their prioritization for implementation as a bilevel, stochastic mathematical program that accounts for: system-wide traffic impacts from reduced capacity from deterioration and construction work zones, and post-improvement increased capacity and speed (a user equilibrium is sought in a lower-level traffic assignment problem); and explicitly accounting for uncertainty in asset state over time due to stochastic evolution of deterioration processes (a Markov decision process -MDP- problem formulation of the upper-level decision process involving probabilistic state transitions due to deterioration);