AutoPort – When AI optimizes port logistics and management

The AutoPort project is at the forefront of transforming Norwegian ports through innovative research on automation and AI, focusing on developing scalable, adaptable solutions that integrate both autonomous systems and human operators. Significant progress has been made through comprehensive reviews of the current drivers and barriers associated with port automation, alongside detailed mapping of three key Norwegian ports: Narvik, Larvik, and Moss. The project conducted successful port visits to Larvik and Moss in 2023 which provided valuable insights and, together with ongoing interviews with port organizations, operators, and technology providers, continue to enhance the project’s understanding and development of the AutoPort use cases. The cooperation with Narvik is also advancing actively. The project continues to ensure stakeholder engagement through interviews and workshops in order to validate and refine models for autonomous port operations. These interactions are instrumental in defining the future roles, functions, and requirements needed for the automation process, ensuring that current practices are aligned with anticipated technological advancements. So far, the project have identified significant operational complexities, bottlenecks, and areas for improvement in yard management, particularly in the manual scheduling and routing of port vehicles (e.g., reach-stackers) for internal transport and the container yard stacking process. Both processes are currently executed without decision support or automation, highlighting substantial potential for enhancement. The project has prioritized addressing these challenges, initially focusing on the scheduling and routing of reach-stackers and exploring the potential and performance of automated guided vehicles (AGVs). A literature review on the topic revealed significant research gaps, prompting a focus on AGV scheduling and routing in container terminals with conflict avoidance. This problem requires generating an optimal plan for AGV operations to minimize the makespan for loading and unloading by determining both scheduling (which AGV handles each container and when) and routing (how containers are transported from their origin to destination efficiently). Over the past year, the team has developed a state-of-the-art mathematical model capable of solving small-scale instances and designed a matheuristic solution method that can handle real-world scenarios. The implementation of these models are nearing completion, with testing set to begin soon. The detailed project’ postdoc report outlines the progress made, emphasizing the potential impact these solutions could have on small port operations. The project has also played a pivotal role in supporting the creation of an Autonomous Port model and advancing the broader development of automation concepts, including evolving roles and new requirements for a Port Operation Centre. This modelling process has been essential for validating the tentative models, aligning the "As-is" situation with input from relevant stakeholders, and refining the vision of the "To-be" scenario concerning future automation processes. The insights gained from these activities have been critical in calibrating the port models and informing the design of the Port Operation Centre, which is envisioned as the nerve centre for future autonomous ports. The Centre could oversee key functions such as cargo flow optimization, plan adjustments, and sensor management, facilitating real-time decision-making, and seamless integration of automated and human-operated systems. This comprehensive approach not only addresses immediate operational challenges but also sets a strategic path for the gradual implementation of advanced technologies thereby paving the way for more efficient, responsive, and resilient port operations in Norway. By focusing on scalable, research-driven solutions, the AutoPort project is positioning Norwegian ports as pioneers in the global shift toward smart, autonomous logistics, driving forward the roadmap for integrating AI, automation, and operational research in small- and medium port environments.

AutoPort Concept will focus on how autonomy and AI can be combined with advanced Operational Research to develop the next generation of autonomous ports. The project will exploit the expert domain knowledge from the partners representing ports, technology suppliers , autonomous ship operators , terminal operator systems, and transport authorities. The AutoPort Concept consists of three main levels: i) the operational level, where Operational Research and AI planning for equipment coordination and scheduling of tasks is used, ii) the tactical level, where management tactics for flow of goods and supervision of the autonomous and manually controlled port processes are developed in an operation centre, and iii) the strategic level, where intermodal transport chains are optimized in a long-term perspective and strategies for implementing autonomy and AI are developed. AutoPort will aim at realizing a long-standing political goal in both Norway and Europe: the modal shift from road to seaborne and rail transport. This shift will enable a more efficient, safer, and more sustainable transport system. Secondly, the project aims at spurring interest in business cases for the Norwegian maritime industry based on the AutoPort Concept. The Concept will be developed in cooperation with Norwegian stakeholders, as well as with public authorities in ports, transportation, and maritime industry to make sure that the concept fits into the Norwegian decision-making framework (e.g., the National Transport Plan), as well as to be used by the industrial partners for their business cases.