State Estimation and Environment Recognition for Articulated Structures Via Proximity Sensors Distributed Over the Whole Body
Kengo Iwao, Hikaru Arita, Kenji Tahara
AI summary
Problem
Kinematics alone cannot accurately estimate the posture of lightweight, low-rigidity articulated robots, especially when their entire body contacts the environment. Existing camera-based SLAM methods are ineffective at close range or during continuous contact.
Approach
The authors distribute proximity sensors across the robot's links to capture environmental point clouds, then extend the discrete-time SLAM model spatially along the link chain to recursively estimate the full-body state and map surroundings simultaneously.
Key results
- Significantly reduced posture estimation errors in simulations
- Simultaneous whole-body state estimation and environmental mapping
- Spatial propagation of state uncertainty minimizes cumulative link errors
- Validated approach for contact-heavy robot navigation scenarios
Why it matters
Enables reliable navigation and manipulation for lightweight articulated and soft robots operating in confined, contact-rich environments where traditional sensors fail.
Abstract
For robots with low rigidity, determining the robot’s state based solely on kinematics is challenging. This is particularly crucial for a robot whose entire body is in contact with the environ- ment, as accurate state estimation is essential for environmental in- teraction. We propose a method for simultaneous articulated robot posture estimation and environmental mapping by integrating data from proximity sensors distributed over the whole body. Our method extends the discrete-time model, typically used for state estimation, to the spatial direction of the articulated structure. The simulations demonstrate that this approach significantly reduces estimation errors.