International Centre to Centre – Machine Intelligence with Robot Swarms for Underground Operations

Vision

Our vision is failure-free operation of millions of kilometres of buried water pipes, enabled by multiple autonomous robots working collaboratively.

Research programme and objectives

The programme is based on active research collaboration with Massachusetts Institute of Technology (USA), Boston University (USA), TU Darmstadt (Germany) linked with the Robotics Institute Germany. The research programme will deploy hybrid machine intelligence approaches, from bio-inspired swarm robotics algorithms, to state-of-the-art statistical machine learning approaches, including deep nets and reinforcement learning, to develop a robust suite of solutions to the problem of operating robustly and efficiently under severe constraints on communications, compute, and power. The following research objectives (RO) are proposed to address the research challenges:

Research Objectives

• To develop a set of performance requirements for a robot swarm operating autonomously in realistic subterranean environments.
• To develop new science for machine intelligence in a robot swarm operating autonomously in these challenging environments to meet the requirements identified in RO1.
• To develop and integrate the best sensing, communication and control methods with a scalable intelligent robot swarm of variable size and complexity to work in underground pipes.
• To apply robust uncertainty quantification methods for trustworthiness of the swarm robotics (from RO2-RO3) operating in subterranean environments.
• To carry out performance validation through simulations and experiments with the proposed robot swarms.

Work areas

• WA1 Mapping and localisation with swarms
• Heterogeneous sensor data
• Swarm control algorithms
• Robots for pipe inspection and maintenance

Core Team

• Professor Kirill Horoshenkov, University of Sheffield (Principal Investigator)
• Professor Simon Tait, University of Sheffield (Co-Investigator)
• Professor Joby Boxall, University of Sheffield (Co-Investigator)
• Professor James Marshall, University of Sheffield (Co-Investigator)
• Professor Lyudmila Mihaylova, University of Sheffield (Co-Investigator)
• Dr Mohamed S. Talamali, University of Sheffield (Co-Investigator)
• Dr Lin Cao, University of Sheffield (Co-Investigator)
• Dr Zhixiang Chen, University of Sheffield (Co-Investigator)

Overseas academic partners

• Professor Roderich Gross, Technical University at Darmstadt, Germany
• Professor Kamal Youcef-Toumi, Massachusetts Institute of Technology, USA
• Alyssa Pierson, Boston University, USA

Industry Partners

• Minicam (robotic equipment manufacturer)
• Opteran (developers of neuromorphic software for robots)
• Nvidia (developers of GUI, API and AI)
• Siemens (developers of data analytics and digital twins)
• Northumbrian Water (water utility)
• Severn Trent Water (water utility)
• Stantec (Engineering Consultants to water utilities)

More information

• Funder: EPSRC (project reference: UKRI1486)
• Value: £1.8 million
• Start date: 1^st May 2026
• End date: 31^st August 2029