PhD Position in Physical AI: Adaptive Foundation Models for Robotics
Overview
The Graduate School of Technical Sciences at Aarhus University in Denmark is offering a three-year PhD position in Physical AI within the Electrical and Computer Engineering programme. The candidate will join the Adaptive & Agentic AI Lab, supervised by Associate Professor Behzad Bozorgtabar and co-supervised by Professor Qi Zhang. The position is available from 1 January 2027 or later.
The project asks how intelligent robots can understand instructions, anticipate the consequences of their actions and adapt reliably when the physical world changes. It connects multimodal perception, reasoning and action with predictive learning and edge intelligence. Research directions include vision-language-action models that link visual observations and language instructions to robot behaviour, world models that predict how the physical world responds to actions and support planning, and adaptation and edge intelligence, covering test-time and continual adaptation, uncertainty-aware decision-making, and efficient inference under latency, memory and energy limits. The exact focus will be shaped with the successful candidate. Work will emphasise new learning algorithms and careful evaluation on public datasets, simulation and, where available, robotic and edge platforms, aiming at publications at leading machine learning, computer vision and robotics conferences.
Requirements
Applicants need a Master's degree of 120 ECTS or equivalent, completed by enrolment, in computer science, electrical or computer engineering, robotics, machine learning or a related field, with strong academic results and foundations in machine learning, linear algebra, probability and optimisation. Strong Python and PyTorch skills, or a comparable framework, are required. Only applicants with substantial hands-on experience implementing, training and evaluating deep-learning models will be considered, and general interest or experience limited to tutorials or pretrained-model interfaces is not enough. Applicants must show research potential through a substantial thesis, research project, code contribution or publication, clearly identifying their own technical contribution.
How to apply
Applications are submitted online through the university, and a project description must be uploaded as a PDF.
Key dates
The deadline is 1 November 2026 at 23:59 CET, and the preferred start date is 1 January 2027.