Robot Kinematics

Joint motion is the motion of a robot achieved by adjusting the angle or position of each joint. Each joint can rotate or move to get the entire robot tool to an expected position.

The linear motion is the motion of the robot tool on a linear path. Unlike joint motion, linear motion is achieved by moving the robot tool along a linear path.

Relative move is a way to describe the target waypoint. Instead of directly describing the target waypoint in space, relative move defines the target waypoint by describing its offset relative to a reference point, such as the robot’s current TCP pose or a reference frame.

Forward kinematics calculates the TCP based on the known joint positions. By substituting joint positions into the robot’s configuration equations (including link lengths, joint zero positions, and rotation directions), the TCP can be uniquely determined.

When robot motion commands are sent as joint positions (JPs), the robot directly follows these JPs to move. As a result, the executed path matches the path planned in Mech-Viz exactly.

Inverse kinematics calculates joint positions based on a given TCP. Because the number of unknowns is greater than the number of knowns, inverse kinematics typically has multiple solutions, which means different joint position combinations may produce the same TCP.

When tool poses are used for robot communication or motion control, the robot controller must solve inverse kinematics internally. This may lead to differences between the actual robot path and the path planned in Mech-Viz, which can result in deviations or collisions.

However, TCP-based communication is still widely used for several reasons.

To avoid inconsistencies between the robot’s actual pose and the pose planned by Mech-Viz due to multiple solutions in inverse kinematics, some robot manufacturers provide axis configuration parameters in robot motion commands.

An axis configuration typically contains four additional parameters: three parameters constrain the angle ranges of the 1st, 4th, and 6th robot axes; one parameter defines the pose type of the 2nd and 3rd axes. By adding these boundary conditions, it can filter out unwanted solutions and ensures that the resulting robot pose matches the pose planned in Mech-Viz.

Based on their location and cause, robot singularities are generally classified into the following types:

In practical industrial applications, the most common singularity occurs in six-axis spherical wrist robots when the 4th and 6th axes approach coaxial alignment during linear motion.

Six-axis spherical wrist robots inherently have singularities in their motion. The following strategies can help reduce the risk of encountering singularities:

Additionally, Mech-Viz can automatically detect potential singularities during path planning. For detailed methods on singularity detection, see Singularity Detection.