End-Point Position Estimation of a Soft Continuum Manipulator Using Embedded Linear Magnetic Encoders

Area of Science:

• Robotics
• Mechatronics
• Sensor Technology

Background:

• Soft continuum robots offer compliant manipulation but lack integrated proprioception.
• External sensors (cameras, optical trackers) are often impractical for deformable robots.
• Developing embedded, non-intrusive sensors is crucial for control and design.

Purpose of the Study:

• To present design considerations for embedded linear magnetic encoders in soft continuum robots.
• To demonstrate practical results of estimating manipulator tip position using these sensors.
• To evaluate the accuracy and repeatability of the proposed sensing method.

Main Methods:

• Embedding three flexible scales with incremental tracks as passive tendons within the robot structure.
• Utilizing six pairs of Hall effect linear sensors to measure relative displacements.
• Estimating robot curvature and tip position from sensor measurements.
• Comparing embedded sensor data with ground truth from an optical tracker system.

Main Results:

• Successful estimation of tip position for a soft continuum manipulator module.
• Achieved average position estimation errors below 2.0 mm.
• Recorded a peak error of 8.7 mm over a motion span of approximately 100 mm.
• Demonstrated the feasibility of embedded magnetic encoders for proprioception in soft robots.

Conclusions:

• Embedded linear magnetic encoders provide a viable solution for proprioception in soft continuum robots.
• The proposed method offers accurate and repeatable tip position estimation without external sensors.
• This technique enhances the design and control capabilities of deformable robotic systems.