Master's thesis: Can Autonomous Vehicles from Different Brands Work Together in a Mine?

Background The mining industry is undergoing a rapid transition towards increased automation, electrification, and digitalisation. Autonomous vehicles and machines are increasingly being used to improve safety, productivity, and sustainability. At the same time, mining operations often consist of mixed machine fleets from several different suppliers, where autonomous and manually operated vehicles need to work together within the same operational area. Despite technological advances, there are currently no established solutions that enable efficient and safe collaboration between autonomous systems from different manufacturers. This limits flexibility, increases costs, and makes large-scale implementation of autonomous mining operations more difficult. Description To enable the autonomous mines of the future, vehicles, machines, and control systems from different suppliers must be able to share information, understand one another’s intentions, and make coordinated decisions in real time. Today’s autonomous mining solutions are often supplier-specific, creating lock-in, limiting scalability, and making efficient collaboration in shared operational areas more difficult. This is particularly important in Autonomous Operation Zones (AOZs), where autonomous haul trucks, loaders, manually operated vehicles, and higher-level control systems need to coexist under shared traffic and safety rules. The master’s thesis will be carried out within the CAVE project — Interoperable Collaboration of Autonomous Vehicles and Machines for Efficient Mining Operations. The project aims to develop and evaluate principles, architectures, and design guidelines for supplier-neutral interoperability between autonomous and manually operated vehicles in mining environments. A central starting point is the evaluation of a broker-based system-of-systems architecture for coordinating autonomous mining transport operations involving multiple suppliers. The aim of the master’s thesis is to develop an initial version of a simulation platform for evaluating architectural principles that allow vehicles and machines from different suppliers to collaborate in realistic mining scenarios. The simulation will be used to test and validate key assumptions in the architecture, such as how shared situational awareness, intentions, traffic rules, prioritisation, and fallback behaviours can be modelled and how they affect productivity, waiting times, and robustness. Possible research questions include: How are productivity, waiting times, and flow affected by different collaboration strategies between autonomous and manual vehicles? Which parameters are most important for robust and safe traffic management in an AOZ? How can interoperability, intentions, and shared situational awareness be modelled in an agent-based or discrete-event simulation? How can broker- or mediator-like coordination be compared with more centralised or local coordination strategies? The expected outcome is an executable simulation model with documented scenarios, an analysis of key parameters, and a report showing how simulation can be used to validate and further develop the CAVE project’s architecture and design principles. The thesis can thereby contribute both to the project’s technical development and to future standardisation of interoperable autonomous transport systems in mining environments. Key Responsibilities The master’s thesis is expected to include the following tasks: Conduct a literature review on interoperability, system-of-systems, autonomous mining transport, mixed traffic, agent-based simulation, and traffic management in defined operational zones. Become familiar with the goals of the CAVE project, its preliminary architecture, and a forthcoming IEEE ISSE 2026 conference paper on broker-based coordination for multi-brand autonomous haulage. Model one or more Autonomous Operation Zones (AOZs) with relevant actors, such as autonomous haul trucks, LHDs/loaders, manually operated vehicles, traffic rules, zones, intersections, loading points, and unloading points. Develop an initial agent-based simulation model in Python, preferably using Mesa. Alternatively, SimPy, SUMO, or AnyLogic may be used depending on the student’s profile, available tools, and the focus of the model. Implement different coordination and collaboration strategies, such as local priority, central traffic management, and broker-/mediator-based coordination based on limited intention and position information. Define and run scenarios in which vehicles from different suppliers need to share information and handle encounters, intersections, loading, queue formation, waiting times, and potential conflicts. Analyse results in terms of productivity, waiting times, resource utilisation, robustness, and sensitivity to communication delays, information quality, or changed prioritisation rules. Document the simulation model, assumptions, scenarios, parameters, and results in a way that can be reused in the continued work of the CAVE project. Summarise conclusions on how simulation can be used to validate architectural principles, identify requirements, and support future standardisation. Qualifications Suitable backgrounds and meritorious skills include: Student enrolled in a Master of Science in Engineering or master’s programme in, for example, computer science, automation, mechatronics, engineering physics, systems engineering, or a related field. Interest in autonomous systems, simulation, traffic management, logistics, or digitalisation of industrial systems. Good programming skills, preferably in Python. Experience of, or interest in, simulation frameworks such as Mesa, SimPy, AnyLogic, or similar. Understanding of modelling, system architecture, optimisation, agent-based models, or discrete-event simulation. Ability to work independently, structure complex problems, and clearly document assumptions and results. Good ability to communicate in spoken and written English. Swedis