Is workspace representation the same as configuration space?

No. Workspace coordinates describe physical locations or poses; configuration coordinates describe the robot's complete arrangement through independent variables.

Can two configurations reach the same point?

Yes. The two-link example shows different elbow positions with the same tip position.

Does a reachable endpoint guarantee a feasible path?

No. Obstacles, joint limits, and other constraints can prevent the required movement.

AI Glossary / Graphs and symbolic AI / Intermediate

Workspace representation

Definition

A workspace representation describes a robot's geometry or relevant locations using coordinates in the same physical space as the objects it manipulates or avoids. In a planar example, these may be horizontal and vertical coordinates; in a three-dimensional setting, they describe spatial positions and possibly orientations. This contrasts with a configuration representation built from the robot's independent joint or pose variables.

What it means

The representation makes physical relationships easy to express: a tool should reach a point, a link should avoid an obstacle, or an object should remain inside a region. However, arbitrary coordinate values may not describe a physically possible robot. Connected links must preserve their lengths and satisfy joint restrictions.
The word workspace also has a narrower usage in robotics: the set of end-effector positions or poses that a particular robot can reach. This entry follows the source's representation meaning while keeping that related reachable-set usage explicit. Neither meaning refers here to a software project folder or a collaborative office environment.

How it works

Choose a coordinate frame and describe the robot and environment within it. Forward kinematics maps a joint configuration to the positions and orientations of robot parts. Inverse kinematics seeks configurations that realize a requested tool position or pose.
A target may have no solution, one solution, or several solutions. Knowing the end-effector position alone may leave the rest of the robot unspecified. Planning must account for link geometry, obstacles, and constraints throughout movement. A curve drawn through free-looking points for the tool does not automatically define a feasible collision-free motion of the whole robot.

Three practical examples

A picking task can specify a target location relative to a worktable, then use kinematics to determine a suitable arm configuration.
A robot simulator can display links and obstacles in one physical coordinate frame for geometric inspection.
A planning lesson can compare Cartesian endpoint coordinates with joint angles to show why two arm postures can reach the same point.

Think of it this way

Think of describing where a person's hand is in a room versus describing the angles of their shoulder and elbow. The first communicates the physical target; the second says more about the posture used to reach it. The analogy is limited because real bodies and robots have different constraints, and an intuitive human pose is not a substitute for a robot's kinematic model.

Worked example

Consider an original ideal planar arm with two links, each one unit long, anchored at the origin. Let theta1 be the first link's angle from the horizontal and theta2 the second joint's angle relative to the first link.
The elbow is at (cos theta1, sin theta1). The tip is at (cos theta1 + cos(theta1 + theta2), sin theta1 + sin(theta1 + theta2)). These workspace coordinates follow from the two joint angles.
At configuration (0 degrees, 90 degrees), the elbow is (1,0) and the tip is (1,1). At configuration (90 degrees, -90 degrees), the elbow is (0,1) and the tip is again (1,1). The endpoint position is identical, but the arm occupies different parts of the plane. The final link orientations also differ, so these are not identical full end-effector poses if orientation is included.
Place a small circular obstacle of radius 0.1 at (1,0). In the ideal thin-link model, the first configuration collides because its elbow lies at the obstacle's center. The second configuration's links run from (0,0) to (0,1) and then to (1,1), each staying at least one unit from that center, so they avoid this obstacle.
Finally, a requested tip position (3,0) is unreachable with these two unit links, even in an obstacle-free plane, because their total reach is at most two units. These calculations distinguish endpoint description, posture, collision, and reachability. They do not yet supply a collision-free path between configurations or a physical safety assessment for a real arm.

Illustrative case study

Fictional case: a training lab draws a straight line from a robot gripper's current position to a target. The line avoids the visible obstacle, but one inverse-kinematics solution swings an elbow through it.
The team keeps the workspace target for communication while checking full-arm configurations along candidate motions. It also records the coordinate frame and whether orientation is required. The revised exercise teaches why a clear target description is necessary but insufficient for motion planning, without claiming that the simplified simulator certifies a real installation.

Common misconceptions and mistakes

Do not assume every point in the surrounding physical space belongs to the robot's reachable workspace. Do not assume an end-effector position uniquely determines the joint configuration. Do not omit orientation when the task requires it, such as inserting an object along a particular direction.
Also distinguish checking a final posture from checking an entire motion. Links can collide during movement even if both endpoints are acceptable. Coordinate frames and units matter: a correctly computed location in the wrong frame is still the wrong target.

Business use

Workspace representations help people specify and inspect robotic tasks in terms of the objects and locations they care about. Configuration representations help software reason about the mechanism that must perform those tasks.
For a robotics prototype, document frames, units, link geometry, joint limits, and the required tool pose. Keep target specification separate from feasibility and collision checking. When reviewing a demonstration, ask whether it shows only a moving tool marker or the full robot geometry. That distinction helps identify what the demonstration actually establishes.

Check your understanding

Why does the obstacle distinguish the two configurations when both tips are at (1,1)?

Answer: The full links and elbow positions differ. Collision depends on the robot's occupied geometry, not only its tip location.

Related terms

Configuration space
Inverse kinematics
Motion planning
Task space
Forward kinematics
End effector
Joint limit
Collision checking

Learn next

Evaluate the same arm at (0 degrees, 0 degrees) and (90 degrees, 0 degrees). The tips are (2,0) and (0,2). Draw both complete arms, then explain why including link positions provides more collision information than marking only their endpoints.

Sources

Primary sources: AIMA original workspace-representation and linkage-constraint context: https://raw.githubusercontent.com/aimacode/aima-glossary/master/sentences.txt ; Lynch and Park, Modern Robotics, task space and workspace: https://modernrobotics.northwestern.edu/nu-gm-book-resource/2-5-task-space-and-workspace/ . The two-link geometry is original.

Previous: Working - Hotelling procedure Next: Wrapper