Add to Chrome
Add to Firefox
Add to EdgeHaptic Sequential Monte Carlo Localization for Quadrupedal Locomotion in Vision-Denied Scenarios
May 04, 2020
Continuous robot operation in extreme scenarios such as underground mines or sewers is difficult because exteroceptive sensors may fail due to fog, darkness, dirt or malfunction. So as to enable autonomous navigation in these kinds of situations, we have developed a type of proprioceptive localization which exploits the foot contacts made by a quadruped robot to localize against a prior map of an environment, without the help of any camera or LIDAR sensor. The proposed method enables the robot to accurately re-localize itself after making a sequence of contact events over a terrain feature. The method is based on Sequential Monte Carlo and can support both 2.5D and 3D prior map representations. We have tested the approach online and onboard the ANYmal quadruped robot in two different scenarios: the traversal of a custom built wooden terrain course and a wall probing and following task. In both scenarios, the robot is able to effectively achieve a localization match and to execute a desired pre-planned path. The method keeps the localization error down to 10\,cm on feature rich terrain by only using its feet, kinematic and inertial sensing.
