Robot basics

Let's configure the robots motion.

Principle of robot motion

Target position

A target position indicates where to move. Target positions are recorded in a robot program and the robot moves to the target positions repeatedly. As position is the primary target for industrial robots to achieve, we can say that the robots perform position control.

For industrial robots, the action to record the target positions is normally called teaching. In UR robots, the target positions are called Waypoint and are taught by using hands in Freedrive mode or jogging from Teach Pendant.

Trajectory

A trajectory defines how to reach the target position. A robot can move linearly or by curve. In general, the robot's trajectory depends on the space which the robot refers to.

In UR robots, the trajectory is specified by Move type on the Teach Pendant. We can say that four Move types are available for UR robots: MoveJ, MoveL, MoveP and MoveC. Descriptions of each move type can be found in the User Manual.

Once the target positions are taught and trajectories are specified in the robot program, the robot can move repeatedly according to the program. For industrial robots, it is crucial to keep reaching the target positions and repetitive tasks are the target positions and repetitive tasks are what the robots can do the best. Therefore, what the robots can do the best. Therefore, repeatability is regarded as one of the most important technical specifications.

Joint Space

A robotic arm consists of links and joints. It is assumed that the links are rigid bodies with fixed lengths. The robot assumes a specific pose based on a set of joint angular positions. In this context, a joint space can be defined as the space representing the robot in the joint coordinate system.

Because the joints are directly controllable, handling the robot in the joint space is straightforward. For example, when moving the robot from its current pose to a target pose, if the trajectory is not important, we simply rotate the joints from their current angles to the target angles. This process is known as joint movement, or MoveJ in Universal Robots (UR) terminology. Joint movement is typically preferred for applications where the target positions are crucial, but the trajectory is not essential, such as:

Pick and place
Injection molding
CNC and machine tending
Packaging and palletizing
Quality inspection (where the robot stops and inspects at each target point)

Linear Space

Humans live in a three-dimensional linear space defined by X, Y, and Z axes, corresponding to the Cartesian coordinate system. We often prefer to see a robot's position and orientation in this linear space, as most tasks are defined this way. Consider, for example, a task to polish the edge line of a desk. The robot needs to move linearly to make a smooth, straight line. In this case, the trajectory should be defined in the linear space, known as linear movement, or MoveL in UR terminology. The following applications typically require linear movement:

Polishing
Gluing, dispensing, and welding
Assembly

Even in applications where joint movement is preferred, as mentioned above, linear movement is often partially needed. For instance, in a pick-and-place application, the robot typically moves to an approach point just above the object, then descends linearly to pick it up. If the robot were to descend using joint movement, there would be a risk of hitting the object.

To move the robot in linear space, it is necessary to convert the robot's pose values in the Cartesian coordinate system to joint positions in the joint space. Conversely, the joint positions must be converted back to the robot's pose to identify its state in the linear space. This conversion process between the two different spaces is called kinematics.