Utilizing Agent-Based Modeling to Evaluate Operational Impacts of an Incident and Possible Alternatives on U.S. Waterways

Abstract

Operations along inland waterways are particularly vulnerable to severe weather events which can create adverse conditions for towboat operators to manage, resulting in increased risk of accidents. However, it is challenging to know or understand the extent to which new regulations or policies on tow operating procedures may reduce impacts from a given scenario. An agent-based model (ABM) representing actual river operating procedures was developed previously to evaluate emergent characteristics and system-wide impacts arising from the interactions between disruptions, towboat operator responses, and regulations. In this work, we validate the effectiveness of the ABM model for improved decision support by comparing it with observed waterway traffic as captured by Automatic Identification System (AIS) data, then apply the model to a case study based on a 2005 incident on the Ohio River. A series of counterfactual scenarios are simulated to better understand potential impacts of the incident on waterway system performance and potential alternative outcomes given different operational responses and regulations. The analysis indicates that imposing tow movement restrictions, such as waiting for a helper towboat, under adverse river conditions can result in a general slowing of the system but with potentially lesser impacts than what would be caused if the system operates as normal and results in an incident that lasts multiple days. The analysis also indicates that the waterway system can return to normal operations within a matter of weeks.