Human-Derived Disturbance Estimation and Compensation (DEC) Method Lends Itself to a Modular Sensorimotor Control in...

Area of Science:

• Robotics
• Biomechanics
• Control Systems Engineering

Background:

• Human posture and movement control are complex, posing challenges for neurological patient rehabilitation.
• Engineering-inspired, model-based approaches are needed to manage this complexity.
• Existing system identification methods are limited to low degrees of freedom (DoF).

Purpose of the Study:

• To investigate if conflict-free movement coordination between sagittal and frontal body planes in robots emerges from physical embodiment.
• To test a heuristic approach for creating human-derived control systems in robots by incrementally increasing complexity.
• To compare robotic control with human control in a shared posture control laboratory setting.

Main Methods:

• Implemented a human-derived control system based on disturbance estimation and compensation (DEC) modules in a 14 DoF robot (Lucy Posturob).
• Tested the hypothesis that mechanical coupling suffices for inter-plane coordination without full-body control.
• Quantitatively characterized inter-plane dynamics using frequency response functions (FRFs) and assessed stability during squatting movements.

Main Results:

• Mechanical coupling in the robot's body successfully coordinated controls between sagittal and frontal planes during intermediate plane movements and balancing.
• Quantitative analysis provided frequency response functions (FRFs) characterizing the interaction dynamics between body planes.
• Postural and control stability were maintained, with emergent inter-segmental coordination observed during squatting movements.

Conclusions:

• Physical embodiment is sufficient for achieving coordinated control between different body planes in robots, simplifying control architectures.
• This approach advances the development of robots capable of complex sensorimotor functions, such as walking.
• Findings support the use of simplified, embodiment-driven control for enhancing rehabilitation technologies for neurological patients.