kinematic model
A kinematic model is a stripped-down geometric description of a robot's body that captures only its shape and how its parts can move, while ignoring weight, force, and effort. A robot arm, for example, is described as a chain of rigid links connected by joints, and the model records how long each link is, how each link is oriented and offset relative to its neighbour, and what kind of motion each joint allows — a swivel that rotates, or a slider that extends. It is the skeleton of the robot written down as numbers, with no muscles attached.
The reason to build one is that it lets you answer pure geometry questions before worrying about anything else. Given the joint settings, where does the hand end up in space, and which way is it pointing? Or, working backward, what joint settings would put the hand exactly where you want it? Because the model leaves out mass and force entirely, it stays simple and exact: the same joint angles always produce the same hand position, like folding a paper ruler the same way twice and getting the same reach.
A kinematic model is usually written in a tidy, standard form so that any program can read it the same way. One common recipe assigns a small set of numbers to each joint and link, so the whole arm becomes a short table. That shared format is what lets one robot's geometry be plugged into planning software, simulators, and controllers without rewriting the math each time.
Before a factory arm ever moves, its kinematic model is loaded into the planner: a table listing each joint's type and each link's length, so the software can compute that bending joint two by 30 degrees will swing the gripper 12 centimetres to the left.
The model is read once and reused for every motion the arm plans.
A kinematic model handles only geometry and motion shape; once forces, mass, and acceleration enter the picture, you need a dynamic model instead.