The objective of this study was to develop practical scheduling solutions for chemical tankers visiting the Port of Houston (PoH). Chemical tanker movements represent approximately 42% of the Houston Ship Channel traffic. Historically, chemical tanker scheduling has been problematic and has resulted in long waiting times for tankers. Scheduling is difficult because chemical tankers carry several liquid cargoes and must visit multiple terminals for loading and unloading. Physical constraints (layout of the port and draft) and commercial constrains (such as terminal and personnel readiness for cargo handling operations, tank cleaning processes, and inspection requirements) create a complex scheduling problem, long waiting times, and unnecessary tanker movements in the port. These problems cause an increase in the business costs for shipowners, risk of collisions and allisions, production of additional air emissions, and decreases in the operating capacity of terminals. The recent expansion decisions for chemical and petrochemical plants in Houston, Texas, will exacerbate the problem. Significant benefits could thus be gained even for small scheduling improvements. Currently, the scheduling practice of loading/unloading activities in the PoH involves primarily the manual and de-centralized use of the "first come, first served" (FCFS) rule, which results in inefficiencies such as long waiting times and poor resource utilization. We propose two mathematical methods to address the tanker scheduling problem in the port: a mixed-integer programming (MIP) method, and a constraint programming (CP) method. The two methods are formulated as open-shop scheduling problems with sequence-dependent post-setup times. MIP yields optimum results that minimize makespan. However, computation time increases significantly as the number of tankers, or the number of terminals, increases. CP achieves better makespan results in a shorter run time, compared to MIP, for medium to large-scale problems including the problem considered in this case study. Overall, the results show that MIP is more suitable for real-time scheduling tools (hourly and daily), whereas CP is the better option for longer-horizon scheduling problems (weekly or monthly). Our models gave good alternative schedules under short optimization run times. Hence, they can afford decision makers sufficient time to complete multiple optimization scenarios and implementation setups.

No abstract

No abstract

Purpose This paper aims to propose a machine learning-based automatic labeling methodology for chemical tanker activities that can be applied to any port with any number of active tankers and the identification of important predictors. The methodology can be applied to any type of activity tracking that is based on automatically generated geospatial data. Design/methodology/approach The proposed methodology uses three machine learning algorithms (artificial neural networks, support vector machines (SVMs) and random forest) along with information fusion (IF)-based sensitivity analysis to classify chemical tanker activities. The data set is split into training and test data based on vessels, with two vessels in the training data and one in the test data set. Important predictors were identified using a receiver operating characteristic comparative approach, and overall variable importance was calculated using IF from the top models. Findings Results show that an SVM model has the best balance between sensitivity and specificity, at 93.5% and 91.4%, respectively. Speed, acceleration and change in the course on the ground for the vessels are identified as the most important predictors for classifying vessel activity. Research limitations/implications The study evaluates the vessel movements waiting between different terminals in the same port, but not their movements between different ports for their tank-cleaning activities. Practical implications The findings in this study can be used by port authorities, shipping companies, vessel operators and other stakeholders for decision support, performance tracking, as well as for automated alerts. Originality/value This analysis makes original contributions to the existing literature by defining and demonstrating a methodology that can automatically label vehicle activity based on location data and identify certain characteristics of the activity by finding important location-based predictors that effectively classify the activity status.

Pl an ni ng an d Sc he du lin g