Challenge 1
Split the mission set into two columns: FIXED, the parts that are the same every round, and VARIABLE, the parts decided per round. Argue about the borderline ones — that argument is the lesson.
The mission 4 min
Plan a mission set that one fixed program cannot cover — and prove it with a number.
Every route you have planned so far has been the same every time you ran it. This mat breaks that. The three launch pads along the bottom look identical and are reached the same way, and which one a round actually needs is not something you can know while writing the program.
This is the real competition condition. A serious round tells you what to do shortly before you run, not weeks earlier. A program that assumes the answer is right some of the time and useless the rest.
The launch site 5 min
Six stations, and their arrangement is the whole point:
- HOME — the start corner, bottom left.
- A, B and C — launch pads 1, 3 and 5, evenly spaced in a row along the bottom.
- D — the launch gantry, bottom right.
- E — the satellite row, alone at the top of the mat.
Look at what that layout says. A, B and C are interchangeable — same line, same approach, same distance apart. Reaching pad 3 instead of pad 1 is the same driving with a different number in it.
And E is expensive. It is the only station away from the bottom edge, so visiting it means a long crossing up and back that no other mission needs. It is the obvious thing to include when you have time and drop when you do not.
The technique — what varies, and what it costs 12 min
Split your mission list into two columns, and be strict about which side things go on.
Fixed — the same in every round. Starting at HOME. Driving along the bottom line. Squaring up at a pad. Delivering to the gantry at D. Returning to HOME.
Variable — decided per round. Which pad. How many pads. Whether the satellite row at E is worth the crossing.
Now count. If a round can call for any one of three pads, and the satellite row is either in or out, that is three times two — six different runs. Write the six out.
Six fixed programs is six ways to load the wrong one. Under pressure, with a scorer waiting, somebody has to find the right file on the Brick and be certain. That is a way to lose a round that has nothing to do with robots — and it gets worse, not better, the more prepared you are.
The alternative is one program that is told which run to do, at the start, by a human pressing a button. The driving is written once. The choice is data.
That is this whole block. Lesson 35 builds the choosing, 36 makes it scale past a screenful, 37 makes the robot remember what it did, 38 lets it do two things at once, and 39 puts them together.
Build it — on paper 10 min
Three artefacts, and lesson 35 is built directly from the third.
One — the two columns. Fixed and variable, every mission on the set placed in one of them. Argue about the borderline ones; that argument is the lesson.
Two — the costed route for one run. As in lesson 25: station order, a time beside every leg and action, totalled against the round. Cost the version with the satellite row and the version without, so you know what E is worth in seconds.
Three — the menu list. The runs your program will offer, in the order they will appear, each with a name short enough to read on the Brick screen at a glance in a noisy hall. Pad 1 beats Launch pad one delivery run, every time.
The program — one shape, one number 12 min
Write the pad run as pseudocode in your My Block names, once. Then read it back and mark every place where the three pads would differ.
There should be very few. Probably one distance, or one count of lines crossed. Everything else — the follower, the squaring, the attachment, the delivery to the gantry — is identical.
That is lesson 14’s rule at a larger scale. Two blocks that differ only by a number should be one block with an input. Three routes that differ only by a number should be one route with a variable — and the menu is just how a person puts a value into that variable before the robot moves.
Sketch the shape of the finished program in three parts, because you will build exactly this over the next six lessons: choose, then run the chosen thing, then report what happened.
Tune it — defend the menu 9 min
Three questions, answered as a team and written down:
- How many entries does your menu need? Fewer is faster to use and less to go wrong. If two entries do nearly the same thing, merge them.
- What is the satellite row worth? You costed it both ways. Points per second decides whether it is a menu entry or a thing you never do.
- What happens if somebody chooses wrongly? The robot will faithfully run the wrong mission. Decide now whether the menu shows a confirmation before it starts.
Run it — the walk-through, three times 6 min
Block 4’s robot at HOME, pushed by hand. Walk the pad 1 run, then the pad 3 run, then the pad 5 run, with somebody reading the pseudocode each time and a third person timing.
Walking all three back to back is what makes the point land: the three walks feel the same, because they are the same, apart from where you stop.
Then walk the satellite crossing on its own and time it. That number decides whether E is ever worth doing.
The journal 4 min
Today’s entry opens Block 5:
- the mat sketch, corrected in colour
- the fixed and variable columns, with every mission placed
- the number of separate programs the set would need without a menu
- the costed route both ways, and what the satellite row at E is worth in seconds
- the menu list — the runs, in order, with names short enough for the Brick screen
- the three walked times, and how much they differed
Nothing in Block 5 needs a new sensor. Everything from here is software: variables, menus, lists and parallel beams. Your robot is already good enough — this block is about making one program do what six programs would have done.