Lesson 04 · 60 minutes
Every Robot Is Different
Measuring your own robot's numbers
Today's goal
Produce two numbers — milliseconds per room and milliseconds per quarter turn — that are true for YOUR robot, and explain why one measurement was not enough.
Ideas in this lesson
- Identical robots are not identical: wheels, motors, batteries, floor
- Measure, do not guess
- Repeat a measurement to see how much it wobbles
- Differential measurement: subtract two runs to cancel out what they share
- Multiply the error to see it: four turns show a mistake four times bigger
Sense → Think → Do
THINK, with a stopwatch. The robot does the same thing every time; you are measuring what 'the same thing' actually means.
How the session runs
- Open the Calibrate tab. Start at Step 0 and watch a launch: it should creep off the line, not chirp.
- Step 1: two runs from the same start line — one second, then three seconds. Measure both in centimetres and add them as one trial. Do it three times.
- Why two runs? Both start from a standstill, so both contain the same speeding-up. Subtract them and the speeding-up cancels out, leaving pure cruising. This trick has a name in science: a differential measurement.
- Look at the spread. If the three trials disagree by a lot, something is inconsistent — find out what before continuing.
- Step 2: the quadruple spin. Four quarter turns should land the arrow back where it started. Read the miss like a clock face.
- Step 3: the drift test. Three seconds straight, then measure how far off the line it ended up, and apply the trim.
- Step 4: victory lap. A square, using YOUR numbers. If it comes home, you are calibrated.
- Name and save your robot. Write the two numbers on tape and stick it to the chassis.
Sample program
The victory lap: repeat four times — drive half a room, quarter turn left. If your numbers are right, the robot ends where it started, facing the way it started.
V 5 W 200 P 4 F 450 L 350 Q
Paste it into the Workshop's Robot language box, then press Text → blocks to see it as blocks — or just send it straight to the robot.
Try this
- Run the same test on a hard floor and on carpet. Which numbers change, and why?
- Swap in a fresh battery pack and re-measure. How much did speed change?
- Predict the distance for a 2.5-second drive, then test your prediction. (Careful: the ramp is in there too.)
- Work out how far the robot loses to speeding up. You have the numbers: how far it went in one second, and how far it would have gone at full speed for one second.
When it goes wrong
| What you see | What it means |
|---|---|
| Trials disagree by more than about 120 ms | Something is loose, the battery is tired, or the surface changed. The app warns you about this. |
| Robot swerves during the very first moment | Acceleration is too sharp — one wheel is breaking traction. Slow the launch on Step 0 first, then re-measure everything. |
Notes for the grown-up
This is the most important lesson in the whole club and the least exciting. Sell it by refusing to accept a maze run later that was not calibrated — the kids who skip today will fail visibly in week 10 and re-learn it the hard way.