Erwan Renaudo

Welcome to my personal pages

I am currently a post-doctoral researcher in robotics and artificial intelligence. My research focus on how we can provide robot efficient learning abilities in order to increase their autonomy in everyday's life. Typical questions are :
- what are the mechanisms allowing a robot to acquire relevant behaviors in a variety of tasks ?
- How can a robot learn from the simplest sensorimotor associations to the most complex behaviors ?
- How does goal-directed and habitual behaviors, highlighted in neuroscience and psychology studies, relate to these learning mechanisms ?

I think the developmental approach is particularly relevant to address this question : providing mechanisms that allows robot to develop incrementally can make them more and more capable, learning behaviors that fit their embodied capabilities. Learning algorithms and architectures should be thinked in the global life of the robot instead of the few hours it is on for the experiment.

During my PhD, I worked at the interface of computational neuroscience and robotics to design a robotic cognitive architecture inspired by the models of goal-directed and habitual behaviors. My manuscript can be found on the Publications page.