Optimization of Vessel Scheduling Problem and Evacuation Decision Making Using Real-Time Location Data, and Their Economic Impacts: A Case Study for the Port of Houston


Dr. Berna Eren Tokgoz, Assistant Professor, Department of Industrial Engineering


Southeast Texas is a high risk hurricane region. This region is likely to be hit by a hurricane on the average of every seven years. The US Coast Guard is responsible for emergency activities and evacuates all the vessels in ports in case of a hurricane threat. The Army Corps of Engineers do the surveying of an evacuation route and start the evacuation. The vessels need to stop loading or unloading of goods, and need to evacuate the port within 2-3 days. In a regular process, the authority that makes the decision for vessel evacuation is the Pilot Association. Due to the highly chaotic environment during an evacuation, there may not be any pilots assigned to the ships. The risk of collisions or other accidents may get higher in this situation. The substances that are transported in a ship have various hazardous impacts during an evacuation process from the port. Transportation of chemicals needs to be prioritized depending on their impacts and locations to make the safest evacuation decision.


The main objective of this research is to analyze the current vessel transits and eliminate the unnecessary transits within the port. Current rules and regulations have enforcements for vessels to report the departure and arrival locations, but these reports are not open to public for research purpose. The best way to reduce the number of unnecessary movements in the ports is to estimate the reason for transit. The current open access data and even commercial solutions do not state the reason for transit. This research draws the reason for transit network in the port to reduce unnecessary movements for chemical tankers. The research team will use GIS and AIS systems to estimate the reason for transit. Root cause analysis will be performed for the reason for transit and a novel algorithm will be introduced for clustering the reason for transit. The research has a perfect timing to have accurate data with the latest US Coast Guard enforcements. The proposed approach will enable verification of the barges in the terminals and estimation of the berth allocation more accurately. This research will lead to future research that will enable new business opportunities with machine learning algorithms. It is possible to validate current vessel behaviors and find new customer recommendations depending on the historic behavior pattern.


This project was funded by the Center for Advances in Port Management at Lamar University


  • B. Cankaya and B. Eren Tokgoz. “Understanding Vehicle Movement Patterns with Machine Learning Algorithms.   INFORMS, Nashville, TN, USA, November, 2016
  • B. Cankaya and B. Eren Tokgoz. “Optimization of Scheduling for Terminals and their Economic Impacts”, Industrial and Systems Engineering Annual Research Conference and Expo, Anaheim, CA, USA, May, 2016
  • B. Cankaya and B. Eren Tokgoz. “A Novel Decision Support Methodology for Port Logistics”, Industrial and Systems Engineering Annual Research Conference and Expo, Anaheim, CA, USA, May, 2016