Faculty of Science, Technology and Medicine
Faculty of Law, Economics and Finance
Faculty of Humanities, Education and Social Sciences
Interdisciplinary Centre for Security, Reliability and Trust
Luxembourg Centre for Systems Biomedicine
Luxembourg Centre for Contemporary and Digital History
Luxembourg Centre for European Law
Luxembourg Centre for Socio-Environmental Systems
The Doctoral School in Sciences and Engineering is happy to invite you to Loïck CHOVET’s defence entitled
Decentralized Autonomy in Planetary Multi-Robot Teams: Trustworthy Task Negotiation, Resilient Communication, and Adaptive Optimization
Supervisor: Assoc. Prof Miguel Angel OLIVARES MENDEZ
Autonomous multi-robot systems (MRS) are envisioned as key enablers for future planetary exploration and in-situ resource utilization (ISRU) missions. However, achieving reliability and scalability in such systems remains challenging due to communication constraints, heterogeneous architectures, and the absence of trustworthy decentralized coordination mechanisms. This dissertation addresses these challenges by developing a cohesive framework for trustworthy decentralized autonomy in planetary MRS, combining resilient communication, economic coordination, and adaptive connectivity maintenance.
First, this dissertation presents REALMS2 (Resilient Exploration and Lunar Mapping System 2), a foundational MRS framework designed for planetary prospection and mapping. Built on ROS 2 and supported by mesh networking, REALMS2 enables robust, decentralized operation across multiple robots in dynamic environments. The system was deployed during the ESA–ESRIC Space Resources Challenge, where it successfully mapped 60% of an 1800 m² lunar-analog environment while maintaining communication under delays and partial blackouts. This real-world validation establishes the technical baseline for the subsequent research on coordination and network resilience.
Building on this foundation, the second part introduces a blockchain-based Make-or-Buy framework that allows robots to autonomously decide whether to execute or outsource tasks through decentralized economic coordination. Using a multi-parameter cost function that incorporates energy, time, and lifespan depreciation, robots negotiate mapping tasks via blockchain-based auctions. Experiments in simulated lunar environments demonstrate a 17–28% global cost reduction compared to random allocation, while ensuring transparency and non-proprietary coordination among heterogeneous agents.
The third part investigates communication resilience in mesh-based multi-robot networks. A comparative study of ROS 2 middlewares—FastDDS, CycloneDDS, and Zenoh—evaluates their performance under dynamic topologies. The findings identify Zenoh as the most efficient middleware, offering lower latency, reduced overhead, and improved reachability, thus enabling scalable decentralized communication for space robotics.
Finally, a connectivity maintenance algorithm, named BackPropagation, is proposed to preserve network integrity during cooperative exploration. By leveraging local communication metrics and digital elevation models, the algorithm dynamically repositions robots to sustain connectivity while minimizing travel cost. Experimental evaluations demonstrate superior performance over traditional force-based approaches in maintaining network cohesion and energy efficiency.
Overall, this dissertation advances the autonomy and resilience of planetary multi-robot systems by uniting field-proven infrastructure, economic decision-making, and adaptive communication strategies. Together, these contributions establish the foundation for Type 4 Machine Economies—ecosystems where autonomous robots act as economically rational, self-organizing agents collaborating in untrusted and distributed environments.