Dr Kenneth Omokhagbo Afebu

Kenneth currently holds a postdoctoral research fellow position in Engineering department where he completed his PhD with Prof. Yang Liu and Dr Evangelos Papatheou as his supervisors. His research examines the nonlinear behaviour of dynamic systems and uses machine learning to characterise and predict how these systems behave over time. This helps improve reliability, stability, durability, performance, and energy efficiency, especially in systems that can switch between multiple operating states. He also develops intelligent models that can predict unknown conditions a system may face during operation. This led to his work using machine learning and drill-bit vibration signals to monitor bit–rock interactions in rotary-percussive drilling and to detect inhomogeneities in downhole rock layers during deep drilling. These models can help guide drilling decisions by supporting the adjustment of load parameters to maintain effective impact responses while reducing the risk of damaging borehole integrity, particularly in transition zones.

As a novel initiative, he has also extended this approach to harness the sensitivity of the nonlinear behaviours exhibited by small-scale robots in direct contact with human tissue for AI-enabled disease diagnosis. This is based on the principle that, just as inhomogeneities occur in downhole rock layers, different disease conditions also create inhomogeneities within affected tissues. Such inhomogeneities can be sensed through the resulting robot–tissue interaction dynamics, thereby providing a non-visual method of soft-tissue examination.

Research highlights:

As a member of the Exeter Small-Scale Robotics Laboratory, Kenneth’s research highlights include:

• Robotic capsules encountering lesions: Investigating the use of AI and the dynamic responses of a self-propelled robotic capsule as it travels through the bowel and encounters lesions, with the aim of detecting malignant transformation.

• Micro-particles in blood channels: Exploring the use of AI and the nonlinear behaviour of microparticles in blood channels, particularly those in close proximity to cancer sites, to identify metastasis and other intravascular diseases.

• Explainable AI and federated learning in medicine: Investigating how explainable AI, combined with federated learning, can provide a framework for collaboration and innovation, enabling health institutions to train more generalisable and reliable AI models collectively without compromising data security.

• AI-driven analysis of geological systems and drilling responses: Using artificial intelligence to characterise geological variability and drilling behaviour from complex system responses, while integrating multimodal geological and operational data to improve interpretation and reduce uncertainty.

Collaborations:

In carrying out his research, Kenneth collaborates closely with Professors and Clinicians in the fields of:

• Dynamics and control,

• Applied mathematics and

• Gastroenterology

Impact and vision:

Kenneth’s research aim is to revolutionise the modalities of carrying out medical examination during disease diagnosis by taking advantage of the nonlinear sensitivity of micro-robots and their ability to access hard-to-reach anatomy. He will rely on the ability of AI to learning complex nonlinear relationships from data, to map the resulting dynamics of the micro-robots to different disease conditions. This way patients can be offered minimally invasive, but yet, highly effective diagnostic procedures which in the long term may become self-administrable. Thus, allowing more people to be tested and reducing the need for initial face-to-face appointments with clinicians.

Before this, Kenneth worked on the development of a smart system for detecting and quantifying gas leaks along gas transmission lines using machine learning and measurable pipe-flow dynamics. His research experience also includes resonance-based non-destructive testing for crack characterisation in additive manufacturing. In addition, his work spans the use of geophysical methods for subsurface investigation, as well as rock core drilling for mineral exploration and groundwater assessment and development.

Kenneth holds a BSc (Hons) degree in Geology from the University of Ibadan, Nigeria, and a Master of Science degree in Applied Geophysics from the same university before coming to the UK for his Master of Science degree in Petroleum and Gas Engineering with distinction at the University of Salford.