FAEN - Maritime and Ports Operations Research

The global shipping industry faces growing pressure to achieve ambitious decarbonization targets. While the development of alternative fuels is promising, their adoption is limited by restricted supply and inadequate port infrastructure. Wind propulsion is an attractive alternative, capable of reducing the reliance on any type of fuel. However, integrating wind-propelled vessels into conventional fleets introduces considerable operational complexity.
This study presents a novel framework for optimizing wind-propelled vessel deployment within heterogeneous fleets. By leveraging unsupervised machine learning techniques, ports are grouped into optimal clusters to facilitate route design, maximizing wind usage and minimizing coastal segments that erode wind propulsion benefits. A permutation-based algorithm identifies optimal routes within and between these clusters, accounting for cargo volume, energy consumption, vessel speeds, and weather conditions. Empirical data from a major car carrier, combined with specific route-based wind energy calculations, support the model's robustness. Preliminary results underscore both the potential and challenges of wind propulsion. While notable energy savings are achieved when wind-propelled vessels operate on long-haul routes, slower operating speeds can disrupt schedules and necessitate innovative approaches such as dynamic rotations and deployment. Despite these complexities, wind-powered ships offer meaningful reductions in emissions, making them a viable path for maritime decarbonization.

The Panama Canal is a vital hub for international shipping, and the reliability of the canal's operations is critical for the efficiency of global trade. As with other maritime bottlenecks, the Panama Canal operators seek a high utilization of the canal, which may require a buffer of queued ships to prevent canal idle time when disruptions occur, denoted Just In Case (JIC). Conversely, ship operators can reduce both emissions and costs by adjusting sailing speed and arriving Just In Time (JIT). In this paper, we seek to identify the optimal balance between JIT and JIC for the Panama Canal, i.e., a vessel pre-booking policy, while considering uncertainty in vessel arrival times. We propose a simulation based optimization model for evaluating different pre-booking policies' impact on total vessel emissions and canal utilization. The objective is to propose and compare different booking policies and observe their consequences in the reduction of emissions while also ensuring high canal utilization. Our preliminary results indicate that emissions from sailing and waiting can be reduced by pre-booking vessels. Simultaneously, pre-booking of all slots, i.e., no vessel queue, decreases canal utilization during disruptions.

In the burgeoning field of autonomous maritime operations, efficiently coordinating and navigating multiple unmanned surface vehicles (USVs) in dynamic environments is a significant challenge. This study presents an enhanced Q-learning algorithm designed to improve pathfinding for multiple USVs in such settings. The algorithm innovates on the traditional Q-learning framework by adjusting the learning rate, Epsilon-greedy strategy, and penalty and reward functions, integrating a collision avoidance mechanism specifically tailored for complex maritime navigation. Extensive simulations across six diverse scenarios-ranging from single to multiple USVs operations in both static and dynamic obstacle environments-demonstrate the algorithm's superior adaptability and efficiency compared to existing methods. Notably, in single USV scenarios, the improved Q-learning algorithm not only plots more direct paths but also reduces computational demands significantly over traditional path planning methods such as the A$^*$ and APF algorithms. In multi-USV scenarios, it demonstrates robust performance, reducing calculation times by an average of 55.51\% compared to SARSA, 49.14\% compared to the original Q-learning, and 45.26\% compared to the Speedy Q-learning approach. These advancements underscore the algorithm's potential to enhance autonomous maritime navigation, laying a strong foundation for future improvements in the safety and efficiency of USV operations.

This study proposes a Decision Support System (DSS) that integrates historical operational data with weather forecasts to optimize vessel heading during DP operations. Utilizing a mixed-integer linear programming (MILP) model, the framework minimizes fuel consumption by identifying optimal headings while accounting for operational constraints. Preliminary results demonstrate a potential 15.13% reduction in energy consumption, translating to 52.37 tonnes of fuel savings. Whilerealistic real-world operations limit full optimization, this research shows the potential of data-driven heading selection in improving OSV fuel efficiency.

This study develops a stochastic optimization algorithm for fleet deployment, addressing a scenario with two types of demand. The primary demand arises from obligations under the Contract of Affreightment (COA), requiring fixed port calls to fulfill a relatively stable, periodic shipment volume. The secondary demand consists of spot cargo shipments, which can be transported using the spare capacity of vessels assigned to the COA commitments. The objective is to maximize risk-adjusted profit while adhering to constraints on fleet capacity, fixed port calls and shipment quantity within a defined period.