Principal Investigator Matthew Wilson
Co-investigators John Leonard , Hordur Johannsson
Project Website http://isquared.mit.edu/research_detail/14/
While much recent work in mobile robotics focuses on developing algorithms for mapping and localization, the neuroscience community is trying to understand how animals and humans map their environment and localize themselves. These mostly independent efforts are highly complementary and combining them should be beneficial for both fields. We propose to explore the connections between state of the art algorithms in this area in robotics and in neuroscience.
In mobile robotics, this problem is called simultaneous localization and mapping (SLAM) and has been the subject of much activity over the last 15 years. In neuroscience, significant progress has recently been made in understanding some of the mechanisms of the brain that deal with ocalization and mapping. We propose to investigate connections between current SLAM ormulations, in particular the pose graph, and neurophysiological models such as grid nd place cells. We also propose to directly investigate the interaction between grid and lace cells in behaving animals through optogenetic manipulation of grid cell spatial firing properties.