Distributed and Embodied Computing In Dynamic Environments
IDLab-DECIDE conducts cutting-edge research on Distributed and Embodied Computing In Dynamic Environments. Our lines spans several related areas of expertise:
- Generative AI for robotic planning & control. We leverage deep generative models for building powerful world models that are foundational in robotic manipulation, and navigation; thereby combining perception, action and learning.
- Cognitive control. We integrate theories from cognitive science and neuroscience that model human behavior and decision making with state-of-the-art machine learning to create more intuitive interfaces and robots that predict and adapt to human behavior.
- Distributed perception and control. We investigate the fusion of data from multiple high-dimensional sensors and develop planning algorithms to enable robust, distributed decision-making in multi-sensor and multi-agent cyberphysical systems.
- HW-SW co-design of AI algorithms. Restrictions on latency and privacy mandate the deployment of advanced sensing and control techniques on edge accelerators. Our research covers novel AI model training procedures, adaptive computational paths, and architectural innovations optimized for emerging hardware accelerator platforms.
Our fundamental research in distributed and embodied computing is applied across diverse sectors, including agriculture, industry and remote sensing.
More information: see our DECIDE team website.
