Reinforcement Learning Swarm Systems for Autonomous Manufacturing in Unstructured Environments (Ref: SF-JO-2026/3)

About the Project

The ‘what’

This PhD will develop reinforcement learning-driven swarm robotic systems that enable multiple robots to collaborate autonomously in unstructured and dynamic manufacturing environments.

The ‘why’

Future manufacturing systems must move beyond fixed, highly structured production lines toward flexible, resilient, and reconfigurable environments capable of handling variability in materials, tasks, and layouts. Traditional single-robot automation struggles in such settings due to limited adaptability and scalability.

Swarm systems—comprising multiple simple agents coordinating through local interactions— offer a promising solution. However, achieving robust, safe, and efficient coordination in realworld manufacturing remains an open challenge. This project addresses a key question: how can reinforcement learning enable scalable, adaptive swarm intelligence for real-world industrial tasks?

Success will support next-generation manufacturing in sectors such as aerospace, construction, and logistics, contributing to productivity, sustainability, and workforce augmentation.

The ‘who’

You will be based at Loughborough University, joining a multidisciplinary research team with expertise in robotics, artificial intelligence, and autonomous systems. Supervision will include leading academics in multi-agent systems and intelligent automation, with opportunities to collaborate across UK partner institutions.

Industry engagement and sponsorship

The project will be developed in collaboration with industrial stakeholders in manufacturing and automation (e.g. system integrators and advanced manufacturing companies). These partners will provide use cases, validation environments, and real-world constraints, with opportunities for industrial placements or secondments.

Aims and objectives

• Develop multi-agent reinforcement learning (MARL) algorithms for cooperative task execution
• Design distributed coordination strategies for swarm systems under uncertainty
• Enable adaptive task allocation and reconfiguration in dynamic environments
• Ensure safe and robust deployment of swarm systems in real-world manufacturing scenarios

Methodology

The project will combine simulation, AI, and robotic experimentation:

• Develop and evaluate multi-agent reinforcement learning frameworks (e.g. decentralised and centralised training approaches)
• Explore emergent behaviours and coordination strategies through learning and local interaction rules
• Integrate perception and environment modelling for operation in unstructured settings
• Use simulation-to-real transfer techniques to bridge the gap between virtual training and physical deployment
• Validate approaches on multi-robot platforms performing representative manufacturing or assembly tasks

Skills and development

You will gain expertise in:

• Reinforcement learning and multi-agent systems
• Robotics and autonomous systems
• Distributed control and optimisation
• Simulation and real-world robotic experimentation

You will also develop key transferable skills in teamwork, problem-solving, and working on complex interdisciplinary challenges.

Career pathways

This PhD will prepare you for careers in:

• Autonomous systems and robotics (industry or academia)
• AI and machine learning engineering
• Advanced manufacturing and digital automation
• Research and development in distributed intelligent systems

Why Loughborough

With links to MIT (Massachusetts Institute of Technology), Loughborough University offers a leading research environment in robotics, intelligent automation, and manufacturing innovation. You will benefit from:

• Access to advanced robotic platforms and experimental facilities
• A strong interdisciplinary and collaborative research culture
• Structured doctoral training and industry engagement
• A supportive environment focused on research excellence and career development

Name of primary supervisor/CDT lead:

John Oyekan j.o.oyekan@lboro.ac.uk

Entry requirements:

Applicants should have:

• A first-class or 2:1 degree in Robotics, Computer Science, AI, Electrical/Electronic Engineering, or a related field
• Strong programming skills (e.g. Python, C++)
• Good foundation in mathematics (e.g. probability, optimisation, linear algebra)

Desirable:

• Experience with reinforcement learning or machine learning
• Familiarity with robotics frameworks (e.g. ROS) or simulation environments
• Exposure to multi-agent systems or control

English language requirements:

Applicants must meet the minimum English language requirements. Further details are available on the International website.

Bench fees required: No

Closing date of advert: 31 October 2026

Start date: 01 February 2027

Full-time/part-time availability: Full-time 3 years

Fee band: 2026/27 Band RB (UK £5,238, International £29,500)

How to Apply:

All applications should be made online. Under Campus, please select Loughborough and select Programme "Electronic, Electrical and Systems Engineering’. Please quote the advertised reference number ‘SF-JO-2026/3’ under the finance section in your application.

Applications must include a personal statement, up-to-date curriculum vitae (CV), details of two referees (one from your highest degree qualification), certified certificates and transcripts for all completed degree programmes, and a reference to the project ‘SF-JO-2026/3’.

Submission of a research proposal is not essential but may strengthen your application.

To avoid delays in processing your application, please ensure that you submit the minimum supporting documents above.

Project search terms:

artificial intelligence, computer science, control systems, data science, engineering other, robotics, manufacturing engineering

Email address Wolfson:

ws.phdadmin@lboro.ac.uk