Embodied Learning and Intelligence for eXploration and Innovative soft Robotics (ELIXIR) Lab

Embodied Intelligence for Adaptive Robotic Systems

At the ELIXIR Lab (Embodied Learning and Intelligence for eXploration and Innovative soft Robotics), we study how physical systems, like robots, can learn, adapt, and make intelligent decisions. By combining artificial intelligence, soft robotics, and embodied cognition, we’re developing robots that can work independently in unpredictable, real-world environments, on Earth and in space.

We focus on embodied AI — the idea that intelligence comes not just from the brain (software), but also from the body (hardware). Our research looks at how learning algorithms and flexible, sensor-packed robot bodies can evolve together to create more capable machines. We use techniques like reinforcement learning, neural networks, and advanced control systems, and we apply them to real robots, not just simulations.

Our work has many practical applications: from robots that assist with daily tasks in homes or hospitals, to flying robots that can handle tools, to automated systems used in factories, on farms, or in space missions. We design and test a wide range of soft and hybrid robots that help us understand how machines can learn, adapt, and stay reliable, even in tough or unexpected conditions.

By rooting intelligence in real-world physical experience, our goal is to connect today’s data-driven AI with tomorrow’s truly autonomous, general-purpose machines. We’re working to expand what robots can do, so they’re not just smart, but also adaptable, resilient, and ready for the real world.