Challenge 1
Sketch the attachment from the side and from above, marking the mouth, the compartments, the mount and where the motor sits, before you pick up a brick.
The mission 4 min
Collect the marine items, keep them told apart, and still have them aboard after the robot spins on the spot.
Read your sorter sentence from lesson 17 aloud before anyone opens a box. Block 2’s arm only had to hold one thing. This mat asks for more, and the difference is where most teams over-engineer.
Port 2 stays empty. It has been reserved since lesson 2 and the Gyro Sensor goes in it next hour. Whatever you build today must leave that port and its mounting space clear — finding out in lesson 19 that the sorter is sitting where the gyro goes costs you both.
Where it has to work 5 min
The four reefs — A and B across the north, C and D across the south — are where the collecting happens, and they are arranged in a square. That means the route between them is mostly corners.
So the attachment has to survive turning. Not a gentle arc: a point turn, where the robot spins about its own centre and everything on board swings sideways. Test that today, at a reef, with the mechanism loaded.
The lighthouse island at E and the octopus at F sit between the reefs and the harbour dock. Whether they are on your route was decided in lesson 17; if they are, check the attachment reaches them too, because they are not shaped like reefs.
The technique — sort with geometry, not code 12 min
“Sorting” sounds like a programming problem and it is almost never solved that way. The cheapest, most reliable sorter is two places to put things.
If items of one kind go into one compartment and items of another kind into a second, the robot has sorted them without a single extra block of code, and nothing can get muddled by a sensor misreading. Compare that with detecting each item and steering it — more failure modes, more tuning, more to go wrong in a round.
Three principles to design against:
- Separate at the moment of collection. The easiest time to decide where something goes is as it comes aboard, because the robot already knows which station it is at.
- Walls beat cleverness. A divider that physically stops two items mixing will never fail. A rule that keeps them apart eventually will.
- Design for the turn. A tray that holds items by gravity alone empties itself on a spin. Something needs to stop them sliding out sideways.
Ask what happens if the robot arrives crooked. It will. An attachment with a wide mouth that funnels items in is worth more than a precise one that only works from dead ahead — and by lesson 23 the route is four legs long, so “dead ahead” is not a safe assumption.
Build it 10 min
Sketch first — side and top — marking the mouth, the compartments, the mount and where the motor sits. Then build.
Keep lesson 10’s interface rules. It fits one way. It locates before it fastens. It does not move the wheels, the Colour Sensor or the Ultrasonic Sensor, because your follower is tuned to where those sit and lesson 21 is about to depend on the geometry as well.
Drive it from port A as before, keeping the motor on the robot rather than on the removable part. And leave port 2 and its mounting space alone.
Program it — extend Collect 12 min
You already have a Collect My Block from lesson 13. Today it either gains an input or gains a sibling, and which one is a genuine design decision.
If your sorter separates items mechanically, Collect probably does not change at all — the compartments do the work and the program never knows the difference. That is the good outcome.
If your design needs the robot to act differently for different items, give Collect an input saying which — and remember lesson 14’s warning that not every block wants one. An input here has to earn its place.
Find the motor numbers by hand as before, read the degrees off the port view, and write them in the journal. Whatever you build, the main canvas should still read as intentions, not motor blocks.
Tune it — the spin test 9 min
Load the attachment as it would be mid-route, then run a full point turn on the spot. Then another, the other way. Then a fast one.
Anything that falls out, shifts compartment or jams is a failure you have just found for free. Fix it mechanically — a lip, a higher wall, a lighter load — and repeat until three spins in a row change nothing.
Then re-run lesson 10’s check that still matters: put the attachment on and off three times and confirm the Colour Sensor still reads the same value over the same patch of mat. If that number moves, the interface has quietly regressed.
Run it 6 min
Drive from HOME to the north-west reef at A, collect, point-turn towards the north-east reef at B, drive across, collect again, and return to HOME with both items still aboard and still in the right compartments.
Two stations and one turn is the smallest honest test of a sorter, because one item can be held by luck and two cannot.
The journal 4 min
Today’s entry:
- the labelled sketch, with compartments and mount marked
- how your design tells items apart — and whether it does so mechanically or in code, with the reason
- the motor numbers, and whether Collect gained an input
- the spin test: what fell out first, and what fixed it
- the sensor check after three swaps
- one sentence on how the built attachment differs from lesson 17’s sorter sentence, and why