Kristian Ceder | Searcharxiv

arXiv2025

Infrastructure-based Autonomous Mobile Robots for Internal Logistics -- Challenges and Future Perspectives

The adoption of Autonomous Mobile Robots (AMRs) for internal logistics is accelerating, with most solutions emphasizing decentralized, onboard intelligence. While AMRs in indoor environments like factories can be supported by infrastructure, involving external sensors and computational resources, such systems remain underexplored in the literature. This paper presents a comprehensive overview of infrastructure-based AMR systems, outlining key opportunities and challenges. To support this, we introduce a reference architecture combining infrastructure-based sensing, on-premise cloud computing, and onboard autonomy. Based on the architecture, we review core technologies for localization, perception, and planning. We demonstrate the approach in a real-world deployment in a heavy-vehicle manufacturing environment and summarize findings from a user experience (UX) evaluation. Our aim is to provide a holistic foundation for future development of scalable, robust, and human-compatible AMR systems in complex industrial environments.

cs.RO↗

arXiv2024

Curriculum Reinforcement Learning for Complex Reward Functions

Reinforcement learning (RL) has emerged as a powerful tool for tackling control problems, but its practical application is often hindered by the complexity arising from intricate reward functions with multiple terms. The reward hypothesis posits that any objective can be encapsulated in a scalar reward function, yet balancing individual, potentially adversarial, reward terms without exploitation remains challenging. To overcome the limitations of traditional RL methods, which often require precise balancing of competing reward terms, we propose a two-stage reward curriculum that first maximizes a simple reward function and then transitions to the full, complex reward. We provide a method based on how well an actor fits a critic to automatically determine the transition point between the two stages. Additionally, we introduce a flexible replay buffer that enables efficient phase transfer by reusing samples from one stage in the next. We evaluate our method on the DeepMind control suite, modified to include an additional constraint term in the reward definitions. We further evaluate our method in a mobile robot scenario with even more competing reward terms. In both settings, our two-stage reward curriculum achieves a substantial improvement in performance compared to a baseline trained without curriculum. Instead of exploiting the constraint term in the reward, it is able to learn policies that balance task completion and constraint satisfaction. Our results demonstrate the potential of two-stage reward curricula for efficient and stable RL in environments with complex rewards, paving the way for more robust and adaptable robotic systems in real-world applications.

cs.LG↗