Moving the body and placing the tool

A wheel’s rotation predicts travel only under assumptions about contact. Slip — open definition occurs when rotation and surface displacement do not match the simple rolling model. Loose soil, obstacles and terrain slope can change traction. A motor encoder can accurately measure wheel rotation while the rover’s location estimate still drifts.

Mobility depends on the whole mechanical system, including suspension, clearance, stability and available torque. An obstacle that a wheel can touch is not necessarily one the vehicle can cross. Ground testing helps characterize interactions, but a surface encountered in flight may differ from the test material.

A manipulator moves a tool through joint motion. Its workspace — open definition describes positions it can reach, subject to orientation and other constraints. A target inside that geometric region may still be unusable because of collision risk, joint limits, uncertain surface shape or insufficient support for the contact force.

On Perseverance, cameras and instruments cooperate with a robotic arm and sample-handling mechanisms. Treat the tool, arm and rover body as one interacting system. A command that is valid for an isolated joint can still create a poor end-to-end operation.

In a fictional level-ground trial, wheel rotation corresponds to 10 meters of ideal rolling travel, but independent imagery estimates 8 meters of actual motion. The difference is 2 meters. Expressed relative to ideal travel, that is 20 percent. This lesson defines the comparison explicitly because slip conventions can vary.

Driving another nominal 10 meters does not guarantee that the error will cancel. The surface may continue to slip or change. A sensible response combines observations with conservative movement rather than integrating wheel rotations as unquestioned truth.

Before placing a tool, check the body pose, arm path, surface orientation and expected forces. The return path matters too: a measurement is not a complete success if the tool cannot be withdrawn into a usable configuration. Telemetry and images should establish what actually happened.

In a fictional sampling choice, a slightly less attractive target with a verified approach may be preferable to a scientifically exciting one beyond the validated workspace. This does not make science secondary; it recognizes that a measurement must be physically achievable to answer the question.