# Week 2 — Differential-drive kinematics

DIS OpenLab · Foundations of Autonomous Systems · v0.1.0-prepilot · 2026-10-15
Track: simulation/computational model; hardware measurement optional and de-energized
License proposal: CC BY 4.0 for original text, pending rights review

## Learning objectives

Learners will calculate and test how left/right wheel angular rates produce body linear and angular velocity; predict straight, in-place, and arc motion; preserve radians, seconds, and metres; and quantify prediction error without hiding model assumptions.

## Model

For wheel radius `r`, wheelbase `L`, and wheel angular rates `ωL`, `ωR`:

```text
v = r(ωL + ωR) / 2
Ω = r(ωR - ωL) / L
```

`v` is body-forward metres/second. `Ω` is yaw radians/second. The model assumes planar motion, rigid geometry, effective rolling radius, and no lateral slip. Those are hypotheses, not guarantees.

## Required resources and safety

- Starter code and calculator/spreadsheet.
- Use the supplied default computational geometry only for the reference exercise: wheel radius 0.0205 m, wheelbase 0.052 m.
- If measuring a chassis, power must be isolated; keep fingers clear of wheel pinch points. Do not lift a powered rover by its wheels.

## Session A — predict before running

Calculate `v` and `Ω` for:

| Scenario | ωL rad/s | ωR rad/s | Predicted v m/s | Predicted Ω rad/s |
|---|---:|---:|---:|---:|
| Straight | 2.0 | 2.0 | | |
| In-place left | -2.0 | 2.0 | | |
| Right arc | 3.0 | 1.5 | | |
| Stop | 0.0 | 0.0 | | |

State the expected path shape before calculating a number. Check signs against the declared coordinate frame.

## Session B — deterministic integration

Use `wheel_targets_to_twist` and `integrate_pose` in a short Python file or interactive session. Integrate each command for 20 steps at 0.1 s. Record start pose, end pose, and whether heading is normalized between -π and +π.

Then run the supplied test suite. Inspect, do not merely quote, the straight-motion and in-place-turn assertions.

Optional Webots activity after native validation is available: run the same commands from a recorded start pose, read simulator pose, and compare displacement and heading. Until that run happens, leave observed columns blank and label them `NOT RUN`.

## Student lab sheet

Configuration:

| Field | Value | Source/measurement method |
|---|---:|---|
| Wheel radius m | | |
| Wheelbase m | | |
| Step duration s | | |
| Number of steps | | |

Results:

| Scenario | Predicted end x m | Predicted end y m | Predicted heading rad | Observed source | Observed values | Difference |
|---|---:|---:|---:|---|---|---|
| Straight | | | | computational / Webots / not run | | |
| In-place | | | | computational / Webots / not run | | |
| Arc | | | | computational / Webots / not run | | |

## Analysis prompts

1. If both wheels are faster by the same factor, which quantities scale?
2. If the measured wheel radius is 5% smaller than the model, what systematic error should appear?
3. Why can a robot turn even when its centre has nearly zero linear velocity?
4. Name two reasons a physical path may disagree with ideal kinematics.

## Evidence to submit

- completed predictions with units and signs;
- executable calculation or clearly shown arithmetic;
- end-pose table with the evidence source named;
- percentage error only where an actual observation exists;
- one limitation and a proposed measurement to test it.

## Instructor notes

Common errors are mixing wheel diameter and radius, using degrees where the function expects radians, dropping the sign of `Ω`, and reporting simulator values as physical observations. Ask learners to estimate order of magnitude before accepting calculator output.

Assessment checkpoint: a correct numeric result without units earns partial evidence only. A negative result with a clear configuration and explanation is valid engineering evidence.

## Low-bandwidth and accessibility option

Provide the formula and table on paper. Teams can share one calculator. Offer tactile or high-contrast axis markers and a text description: “the rover’s forward axis points from axle centre toward the front; positive yaw rotates the front toward the rover’s left.”

## Exit ticket

Explain in two lines how increasing wheelbase changes angular velocity for the same wheel-speed difference, and name the measurement needed before applying the model to hardware.
