Robotic Torso Position Sensing via Sensor-Encoder Fusion

The patent combines a sensor (optical, magnetic, or capacitive) with an encoder to create an integrated position measurement system. The sensor detects position while the encoder provides reference signals, and their outputs are processed together by a microcontroller to achieve high-accuracy position feedback without requiring separate independent systems.

The measurement system is designed to work with multiple types of sensors (optical, magnetic, capacitive) and can be applied to various robotic mechanisms. The same basic architecture handles both absolute position detection and incremental movement measurement, providing universal functionality across different operating conditions.

The system performs preliminary calibration procedures during startup to establish reference positions and scale factors before normal operation begins. The microcontroller stores calibration data in non-volatile memory, so when the robot is powered back on after being off, the previously stored reference information is retrieved and used to immediately restore accurate position feedback without requiring re-calibration.

The system continuously monitors position feedback from both the sensor and encoder, comparing their readings to detect discrepancies. When the robot is powered off and on again, the feedback loop uses stored reference data to verify and correct position information, ensuring reliability even after power interruptions.

The calibration system is designed to be self-performing through automated microcontroller control. The system automatically moves the robotic mechanism through calibration positions, collects sensor and encoder data, calculates calibration parameters, and stores them in memory without requiring manual intervention. This eliminates dangerous manual calibration procedures while maintaining high measurement accuracy.

Calibration is performed automatically during the initial startup sequence before the robot begins normal operation. The system completes all necessary calibration procedures, including position reference establishment and scale factor calculation, before allowing user control, thereby ensuring accuracy is built-in before any potentially hazardous operations occur.

The system dynamically determines the actual rollout distance parameter through automated measurement during calibration, rather than relying on theoretical calculations from belt and pulley dimensions. The microcontroller measures the actual distance traveled for each encoder revolution and uses this empirically determined parameter for all subsequent position calculations, compensating for manufacturing tolerances in the belt and pulley components.

The system uses feedback from both the encoder and sensor to continuously verify and correct position measurements. By comparing the expected position (based on encoder counts and theoretical rollout distance) with the actual measured position (from the sensor), the system can detect and compensate for rollout distance variations caused by manufacturing tolerances, belt stretch, or pulley alignment issues.