Robots That Don't Get Stuck: A Master-Slave Control System

This patent describes a robot control system where a human-operated 'master' robot guides a 'slave' robot, preventing either from getting into awkward, uncontrollable positions by carefully planning the master robot's movements.

What it covers

This system connects a master robot, which is a haptic device controlled by a human operator, to a slave robot that performs tasks. It uses sensors to measure the movements of both robots (master displacement sensor, slave displacement sensor) and actuators to control their forces (master actuator, slave actuator). A 'master target displacement calculating device' takes the slave robot's movement and maps it to a desired movement for the master robot. Crucially, this mapping is designed to avoid any 'singular configurations' for the master robot, which are positions where the robot can get stuck or lose control. The system also uses a 'slave target driving force calculating device' to control the slave robot based on the force the operator applies to the master robot. This setup means the operator feels the slave robot's actions without needing an extra sensor on the slave robot to measure forces it applies to its surroundings, and it solves the problem of robots getting stuck for both the master and slave robots. For example, a surgeon using a haptic controller (master robot) to perform delicate surgery with a robotic arm (slave robot) would experience smooth, precise control without the robotic arm ever locking up.

What it doesn't cover

• —Does not cover master-slave systems that require a separate working force sensor on the slave robot to measure forces applied to the environment (Claim 1).
• —Does not cover systems where the master robot's movement mapping allows it to enter a singular configuration, a position where it loses control (Claim 1).
• —Does not cover systems where the operator feels the master robot's own internal movements or dynamics, rather than just the slave robot's actions (Claim 1).
• —Does not cover systems where the calculation for the master robot's inverse kinematics needs complex, iterative numerical methods to find a solution (Claim 3).
• —Does not cover systems where the master robot has more than six degrees of freedom, or where its rotational joints do not cross at a single point (Claim 4).

The clever bit

The truly clever part is the predefined mapping for the master robot's target movements. This mapping is specifically designed to *exclude* any singular configurations, meaning the master robot is guided away from positions where it might get stuck or lose control, even before it reaches them. This simplifies the control system and improves reliability.

Why it matters

This patent addresses a significant challenge in robotics: preventing 'singular configurations' where robots can become uncontrollable. By designing the control system to avoid these problematic positions from the start, it makes master-slave robotic systems more reliable and safer. This is particularly important for applications requiring high precision and safety, such as surgical robotics or remote manipulation in hazardous environments. It allows operators to maintain intuitive control without worrying about the robot suddenly losing its ability to move freely.

Real-world examples

• 1.Surgical robots with haptic feedback
• 2.Remote manipulation arms for hazardous materials
• 3.Telepresence robots for exploration
• 4.Industrial robots controlled by human operators