The Vrije Universiteit Brussel (VUB) invites applications for a fully funded PhD position in AI-based Compression and Information Theory for Multi-Agent Perception, hosted within the Faculty of Engineering, Department Electronics and Informatics (ETRO). VUB is a leading Dutch-speaking university in Brussels, renowned for its commitment to freedom, equality, and solidarity, and is a member of the EUTOPIA alliance of European universities. The university offers a vibrant, multicultural environment with a strong emphasis on open research and innovation.

This PhD position is part of the prestigious European Research Council (ERC) Consolidator Grant project “Reinventing Multiterminal Coding for Intelligent Machines (IONIAN).” The research focuses on advancing compression and communication systems for autonomous vehicles, smart transportation systems, and unmanned aerial vehicles. As data volumes from intelligent machines (such as video and point cloud signals) continue to grow, traditional storage and communication methods are increasingly inadequate. The IONIAN project aims to disrupt these systems by developing a novel framework grounded in interpretable and explainable AI.

The successful candidate will pioneer network information theory concepts and their practical deployment in autonomous systems, including ground and aerial vehicles and mobile robots. Key responsibilities include formulating and solving multiterminal information theory problems using modern machine learning techniques, designing and analyzing multiterminal data compression and communication architectures for cooperative autonomous machines, and developing semantic-aware compression methods. The research will involve deriving new rate–(task-)distortion–reliability bounds, designing adaptive codecs that prioritize safety-critical bits, and integrating information theory with AI models for cooperative perception. Results are expected to be published in top-tier conferences and journals.

Applicants should have a Master’s degree (or equivalent) in Electrical Engineering, Computer Science, Applied Mathematics, or a related field. Essential qualifications include a strong foundation in information theory, signal processing, or communications; proficiency in machine learning techniques (ideally deep learning); programming skills in Python, MATLAB, or C++ (including frameworks like PyTorch); analytical thinking and mathematical rigor; experience with experimental or simulation-based research; excellent English communication skills; and the ability to work both independently and collaboratively. Desirable qualifications include prior experience with multiterminal source coding, network information theory, distributed optimization, publications in reputable conferences or journals, interest in robotics, autonomous vehicles, or embedded systems, and familiarity with explainable or interpretable AI techniques. Non-EEA nationals must meet permit and residence requirements for VUB.

The position offers a full-time PhD scholarship for 12 months, extendable up to 48 months based on positive evaluation. The grant is linked to government scales and includes numerous benefits: telework allowance or internet fee, cost-free hospitalisation insurance, full reimbursement of public transport commute, campus meal discounts, sports facilities, nursery discounts, and access to learning platforms. The workplace is open, diverse, and supportive, with a strong focus on work-life balance and career development.

To apply, submit your application online via jobs.vub.be by 2026-04-01. Required documents include a cover letter, CV, academic transcripts, research statement (optional), references, and diploma (not required for VUB alumni). The selection process consists of an application review followed by an interview. For questions about the job content, contact Professor Nikos Deligiannis at [email protected].

For more information about the university, its campuses, benefits, and strategic goals, visit jobs.vub.be. To learn more about the EUTOPIA alliance, visit eutopia-university.eu.