This PhD will develop computational methods for intelligent robotic inspection and defect detection and triage. The central idea is to turn multi-view robotic image acquisition into a view-consistent 3D digital surface representation that can support defect localisation, confidence estimation and decisions about when additional sensing or detailed metrology is needed.
The project will focus on modern computer vision and 3D rendering approaches, including multi-view reconstruction, 3D Gaussian splatting, neural rendering and machine-learning-based anomaly detection. These methods will help separate genuine defects from reflections, glare and surface-finish variations, moving beyond independent 2D inspection towards traceable 3D inspection records.
The student will work within Heriot-Watt’s Institute of Sensors, Signals and Systems (ISSS), at the intersection of the computational imaging and robotics teams, and will collaborate with Renishaw on realistic industrial inspection challenges. The project is suited to students interested in robotics, electrical and electronic engineering, computer science, software AI, machine learning, 3D vision or computational imaging.
Software Needs and Skills
Strong programming ability is important, ideally in Python and modern scientific or machine-learning libraries. Experience or interest in computer vision, 3D geometry, Gaussian splatting, neural fields, SLAM, robotic perception and/or experimental validation would be valuable.
Funding Notes
This studentship is intended for applicants eligible for UK Home fee status. Funding will cover UK Home tuition fees and includes a tax-free stipend at the standard Heriot-Watt/UKRI rate, subject to final confirmation. The project will also include support for research equipment/consumables and travel, including conference attendance and collaboration activities where appropriate.
How to Apply / Informal Enquiries
Applicants should have, or expect to obtain, a strong degree in a relevant engineering, computing, robotics, physical sciences or mathematical discipline. Interested candidates are encouraged to prepare a CV, academic transcripts and a short statement describing their interest in computational 3D inspection, robotics or computer vision.
Contact emails: a.halimi@hw.ac.uk, M.Koskinopoulou@hw.ac.uk
More information regarding the groups can be accessed in: https://sites.google.com/site/abderrahimhalimi/home and https://kskmar.github.io/arm2lab/
Funding Notes
This PhD studentship is in collaboration with the industrial partner Renishaw. The project focuses on automated inspection and intelligent defect triage for reflective and transparent surfaces, using methods from robotics, multi-view 3D reconstruction and rendering, 3D Gaussian splatting, machine learning, and uncertainty-aware decision-making. Funding is available for UK Home tuition fees and a tax-free stipend, with additional support for equipment, consumables and travel. The expected start date is before the end of January 2027.
