How Exoskeletons Stay Safe by Detecting Faulty Sensors

This patent describes a system for exoskeletons that detects when a sensor might be faulty by comparing its readings with other sensors, then switches to a backup and activates a safety mode to prevent harm.

What it covers

This patent describes an exoskeleton with a system to ensure safe operation by detecting sensor problems. It works by comparing the data from a user-operated force moment sensor with at least one torque sensor located in a joint mechanism. If the system identifies a discrepancy between these sensor readings, it can then recruit a substitute sensor from a backup group. After identifying the issue and recruiting a new sensor, a controller activates a specific safety mode and takes a remedial measure to keep the exoskeleton user safe. For example, if a torque sensor in an elbow joint reports zero force while the user's hand sensor shows a strong grip, the system might detect a discrepancy, switch to a backup torque sensor, and then reduce the arm's power to prevent uncontrolled movement.

What it doesn't cover

• —Does not cover exoskeleton safety systems that do not compare output data from at least two different types of sensors, like a force moment sensor and a torque sensor.
• —Does not cover safety systems that only detect sensor failure without recruiting a substitute sensor from a predefined group.
• —Does not cover exoskeletons that implement safety measures without first identifying a discrepancy between specific sensor types.
• —Does not cover general mechanical failures in an exoskeleton that are not related to sensor output discrepancies.
• —Does not cover safety protocols that do not involve executing a 'remedial measure' associated with a 'safety mode'.

The clever bit

The novelty lies in the system's ability to actively compare different types of sensor data (like user-applied force and joint torque) to spot inconsistencies, then dynamically switch to a backup sensor and engage a safety protocol, rather than just relying on a single sensor's self-diagnosis.

Why it matters

Exoskeletons are complex machines that directly interact with human bodies, making safety paramount. A malfunction, especially a sensor failure, could lead to serious injury for the user. This system aims to build trust in exoskeleton technology by providing a robust way to detect and respond to sensor errors, which is critical for their widespread adoption in medical, industrial, and military applications.

Real-world examples

• 1.Medical rehabilitation exoskeletons
• 2.Industrial assist exoskeletons
• 3.Military load-carrying exoskeletons
• 4.Powered prosthetic limbs