Postdoc Efficient quadrupedal locomotion for natural environment exploration

Global warming and pollution are threatening the survival of one million over the eight million species in forests and oceans on the planet. The EU answer to these challenges is the European Green Deal. Among these policies, a prominent role is given to the preservation of ecosystems by increasing the coverage of protected biodiversity-rich land and sea areas building on the Natura 2000 Network (N2000N). Stretching over 18% of the EU’s land area and almost 6% of its marine territory, N2000N is the largest coordinated network of protected areas in the world. The land and sea coverage will be increased up to 30% within 2030. Today human operators are the only option to perform the monitoring of such a large area. This is because their physical intelligence allows them to move for hours in wild unstructured environments. The artificial alternative is robotics. However, nowadays robots hardly leave laboratories and factories because they are not robust and efficient to survive in the real world. The Natural Intelligence (NI) project aims to serve the European Green Deal via monitoring the natural habitats of N2000N with quadrupedal robots able to effectively move in dunes, grasslands, forests, and alpine terrains. NI robots will be empowered by Natural Intelligence, emerging by the interaction of environment, body and mind, leveraging on the fusion of physical and cognitive crucial enablers.

Your role will be to lead the development of the motor intelligence of the NI quadruped, which can exploit the physical intelligence to reach unmatched performance in terms of efficiency and robustness of locomotion. For example, we think of having the quadruped jumping forward for a long distance and land on an unstructured ground without falling or breaking. Think for example of a mountain goat jumping from one rock to another. The algorithms that you will develop will implement precise leg motions, while at the same time preserving the physical intelligence of the quadruped. In analogy to how the central nervous system controls the animal body, we aim at making this possible by integrating a-priori information coming from models and learning strategies.

You will collaborate with researchers from the NI consortium (University of Pisa, ETH, Imperial College of London, CSIC, Kingston University, ISPRA), with the goal of reaching full validation of the proposed algorithms in a real natural environment before the end of the project.