Traversing Visually-Similar Physically-Different Obstacles with DeforMoBot, a Bio-Inspired Deformable Mobile Robot

Barry W. Mulvey, Thilina Dulantha Lalitharatne, Thrishantha Nanayakkara
Imperial College London

Towards Autonomous Robotic Systems (TAROS 2023)
Paper Email Linkedin

Abstract

Many animals have evolved to be able to change their natural shape in order to fit through narrow gaps and spaces. This is generally not possible with rigid robots. In this work, a bio-inspired deformable robot is tested attempting to traverse obstacles of various masses (and therefore, movabilities). The robot uses a shape-adjustment algorithm so that it can traverse obstacles rather than having to circumnavigate them. Our results demonstrate the inverse relationship of the robot’s degree of deformation and its ability to push against compliant obstacles in its path.

Video Demonstration

Poster

Reference

B. W. Mulvey, T. D. Lalitharatne, and T. Nanayakkara, “DeforMoBot: A bio-inspired deformable mobile robot for navigation among obstacles,” IEEE Robotics and Automation Letters, vol. 8, no. 6, pp. 3828-3835, June 2023, doi: 10.1109/LRA.2023.3273393.

BibTeX of RA-L Paper

@article{mulvey2023deformobot,
      author={Mulvey, Barry William and Lalitharatne, Thilina Dulantha and Nanayakkara, Thrishantha},
      journal={IEEE Robotics and Automation Letters}, 
      title={{DeforMoBot}: A Bio-Inspired Deformable Mobile Robot for Navigation Among Obstacles}, 
      year={2023},
      volume={8},
      number={6},
      pages={3828-3835},
      doi={10.1109/LRA.2023.3273393}}