Develop a shared-autonomy teleoperation framework built on tele-visual-tactile fusion for contact-rich nuclear decommissioning tasks.
- Lead Supervisor: Dr Ziwei Wang School of Engineering, Lancaster University
- Second Supervisor: To be confirmed
- Industry Partner: TG0 Ltd (Confirmed - formal agreement in progress)
- Project Start: TBC
- Target Background: 1st or 2:1 / Master's in Robotics, Control Engineering, Mechatronics or Human-Robot Interaction. C++/Python and ROS/ROS2 desirable.
- Industrial Funding: Indicative SME-level support (Sought)
- Advert Close Date: ASAP
- Programme: 4 year Engineering Doctorate (EngD) with industry placement
Tele-visual-tactile fusion for safety critical teleoperation.
The safe decommissioning of key nuclear assets depends on remote robotic systems capable of high-precision manipulation. While visual feedback remains a primary interface, its utility rapidly becomes degraded or wholly insufficient during contact-rich phases, such as valve interaction or seal engagement, where occlusion and limited focal range impede operator judgement.
Under these conditions, conventional teleoperation becomes acutely fragile, increasing cognitive workload and the risk of task failure.
This EngD project will develop a shared-autonomy teleoperation framework built on tele-visual-tactile fusion for contact-rich nuclear decommissioning tasks. The project is anchored by TG0's tactile sensing capabilities (the Linkerbot platform) and Lancaster University's expertise in robotic teleoperation and intelligent communications.
The expected outcome is a new class of operator-centred shared-autonomy methods that improve right-first-time task execution, reduce unsafe force application in visually ambiguous conditions and strengthen the resilience of remote nuclear intervention.
Aims and objectives
- Development of tactile-rich representations for contact-led nuclear tasks. Establish a high-fidelity tactile representation using the Linkerbot dexterous hands, identifying and modelling operationally significant regimes including free motion, exploratory contact, stable engagement and impending slip.
- Design of a tele-visuotactile shared control framework. Formulate a game-theory-based decision mechanism that modulates the balance of human and robot authority, with a force-guided attention fusion module that adaptively weights visual and tactile features without manual annotation.
- Experimental validation via laboratory mock-ups and industrial integration. Evaluate the system through experiments replicating the constraints of nuclear decommissioning, with metrics for peak contact forces, task completion efficiency and operator cognitive load (NASA-TLX).
Alignment to STAND-UP impact targets
- >50% reduction in overall build or decommissioning process time
- >40% reduction in maintenance time
- >30% reduction in person hours on builds (not applicable)
Apply for this project
Contact the lead supervisor or programme team to discuss your interest. Full application instructions are on the How to Apply page.
