Funding
Funded
Learning-based Adaptation for Unmanned Autonomous Vehicles using Visual Foundation Models
School of Electronics, Electrical Engineering and Computer Science | PHD
Applications are now CLOSED
Reference Number
EEECS/2025/MV1
Application Deadline
28 February 2025
Start Date
1 October 2025
Overview
Unmanned Autonomous vehicles (UAVs) have been extensively applied for many practical applications such as nuclear facility inspection, construction, agriculture, security and defence. In many of these scenarios, these are complex situations where the normal operation of the UAVs is close to be disrupted, for example: (i) Perception in complex urban environments, (ii) Perception in adverse weather and poor lighting conditions; (iii) Perception under (partial) occlusions; (iv) Real-time decision-making and motion planning especially in uncertain situations; (v) Self-assessment of perception and localisation systems. Unfortunately, at the moment there are very limited solutions that can guarantee safety at runtime while an UAV explores, especially for challenging environment mentioned above. It is thus necessary to develop a new theory of AI/Machine Learning-based safety critical control for UAVs to guarantee resilience and safety against errors, uncertainties and disturbances. In this project, a new theory of learning (AI/machine learning)-based control adaptation will be developed to mitigate or even eliminate the effects of the errors and disturbances during autonomous operations. A visual foundation model module will be added to explain the environments, which will be used to adapt the control system to guarantee that the UAV will always operate within the designed safe zones during manoeuvring. Meanwhile, learning algorithms (meta learning or reinforcement learning) will be explored to model the uncertainty sources and integrate them within the control system. This will enhance the safety and precision of the perception and navigation of the UAVs.
Unmanned Autonomous vehicles (UAVs) have been extensively applied for many practical applications that are either too dangerous or unsuitable for humans such as environmental monitoring, security surveillance, and search-and-rescue. In many of these scenarios, these are complex situations where the normal operation of the UAVs is close to be disrupted, for example: (i) Perception in complex urban environments, (ii) Perception in adverse weather and poor lighting conditions; (iii) Perception under (partial) occlusions; (iv) Real-time decision-making and motion planning especially in uncertain situations; (v) Self-assessment of perception and localisation systems. Hence, it is imperative to guarantee that both UAV and humans in the surrounding are always safe during operation even when facing unforeseen and unpredictable events.
The objectives of the project include:
[1] To design a learning (AI/Machine learning)-based control framework that ensures the UAVs will always operate within the designed safe zones during manoeuvring.
[2] To design visual foundation models for environment explanation and adaptation.
[3] To develop learning algorithms using meta-learning and/or reinforcement learning concepts, in order to enhance the precision of the safety critical control.
[4] To implement and validate the proposed algorithms in computer simulation platforms: Gazebo platform.
[5] To implement the developed algorithms in a UAV platform, i.e., HuskyA200 AGV robot, which is currently available in our Lab.
The research on ‘learning-based adaption using AI/machine learning’ and ‘unmanned autonomous vehicles’ are hot research topics at the moment. By completing the objectives, the student will obtain multiple skills (i.e., AI/machine learning, hand-on experiments, and programming skills) as well as strong knowledge in robotics, virtual reality and machine learning. The successful student will join our vibrant and world-leading research group and work with the world-class state-of-the-art facilities in our lab, including a xArm robot arm, a Husky A200 AGV, an underwater robot Bluerov2, 2 PAWs quadrupedal robot, a bipedal robot and a robotic laser welding cells to improve their hand-on skills.
Funding Information
To be eligible for consideration for a Home DfE or EPSRC Studentship (covering tuition fees and maintenance stipend of approx. £19,237 per annum), a candidate must satisfy all the eligibility criteria based on nationality, residency and academic qualifications.
To be classed as a Home student, candidates must meet the following criteria and the associated residency requirements:
• Be a UK National,
or • Have settled status,
or • Have pre-settled status,
or • Have indefinite leave to remain or enter the UK.
Candidates from ROI may also qualify for Home student funding.
Previous PhD study MAY make you ineligible to be considered for funding.
Please note that other terms and conditions also apply.
Please note that any available PhD studentships will be allocated on a competitive basis across a number of projects currently being advertised by the School.
A small number of international awards will be available for allocation across the School. An international award is not guaranteed to be available for this project, and competition across the School for these awards will be highly competitive.
Academic Requirements:
The minimum academic requirement for admission is normally an Upper Second Class Honours degree from a UK or ROI Higher Education provider in a relevant discipline, or an equivalent qualification acceptable to the University.
Project Summary
Supervisor
Dr Mien Van
Mode of Study
Full-time: 3 or 3.5 years
Funding Body
Funding TBC