Feeling for Things

Try It — Walking in the Dark

Imagine the power is out and you are trying to cross a room you know well. Your eyes are useless. So you put a hand out in front of you and start walking.

Actually do it. Stand a few steps from a wall, close your eyes, put one hand out, and walk slowly toward it until you touch it.

You did not check once and hope. What were you doing the whole time you were walking?

Sort your sensors

Get your kit’s sensors out on the table — touch, color, gyro, and distance if your kit has one. Without looking anything up, sort them into two piles. You decide the rule.

Learn It — Keep Checking Until

A sensor turns something physical — a press, a color, a distance — into something your program can read.

True or false

The touch sensor’s yes-or-no question comes as a six-sided block — a hexagon. Hexagons do not do anything on their own. Each one is a question with exactly two possible answers: true or false.

Not pressed → the hexagon answers false. Pressed → it answers true. Nothing in between — a light switch, exactly like your sorting rule.

Why the wait block cannot help you here

Every time you needed the robot to wait, you used wait 2 seconds. That will not work now. A sleeping program cannot notice anything. If you tell it to wait two seconds and the sensor is pressed after half a second, the robot never knows.

You need it to keep checking — over and over, as fast as it can. Hexagons snap into two Control blocks built for exactly that:

Wait until asks its question as fast as it can, thousands of times a second, and does not let the next block run until the answer is true. It is your hand-out-in-the-dark, as a block.

Repeat until is the same idea with a mouth: everything held inside its jaws runs over and over, as long as the answer is still false. The moment the answer turns true, the robot falls out of the jaws and moves on to whatever comes after.

Drive until bump

Put it together. Read it out loud: “start driving — keep checking until the sensor is pressed — then stop.”

Comparing numbers

Hexagons are not only about sensors. The green Operators shelf has three comparison hexagons, and they turn any pair of numbers into a true-or-false question:

Today your question is is pressed? and you will not need these. But hold on to them — in Project 11 the loop stops watching a switch and starts watching a number, and these hexagons are how you ask about numbers.

One more: the green not hexagon flips an answer. not (1 is pressed?) is true exactly when the sensor is not pressed. You will want it the moment you need “keep going while nothing is touching.”

Do It — Drive Until Bump

Step 1 — Mount the sensor and read it live

Mount your touch sensor on the front of the robot so it hits things before anything else does. Plug it into sensor port 1 — the numbered ports on the bottom edge are the sensor ports.

On the brick, open Port View and arrow over to port 1. Press the sensor with your finger and watch what it reports.

• My sensor changes when I press it, and I watched it happen before writing any blocks

Step 2 — Drive Until Bump — in your hands first

New project called Drive Until Bump. Plan it as pseudocode first, then build the start-moving / wait-until / stop-moving stack from Learn It.

• Wheels stop when I press the sensor with my finger
• It works on the floor against a wall or a book

Step 3 — Flip the condition and predict

Wrap the hexagon in a green not: wait until not (1 is pressed?). Before you run it, write down what you think will happen.

Now run it. Then run it again while holding the sensor pressed the whole time.

What actually happened, both times? Explain why, using the words “true” and “false”.

Take the not back out.

Step 4 — Bump and Go Home

Here is what makes a touch sensor genuinely useful: a wall is a place your robot can always find.

New project. Mount or turn the sensor so it faces backward — then drive backward until the sensor presses against the starting box wall, and drive forward a set distance into the box and stop.

• My robot finds the wall and ends up in the same place every time

Step 5 — Add a stop and settle

A robot that has been driving is still rocking when the checking ends. Momentum does not care about your program. Put a short wait 0.5 seconds between the stop and whatever comes next, and compare runs with and without it.

Step 6 — Wait for a press before starting

One more use for the same idea. Put a wait until (1 is pressed?) block at the very top of a mission program, before anything moves.

Now your robot waits for a tap on its sensor instead of driving off the second you press Run — which is exactly how you want to start a match.

• My robot waits for a press before it moves

That wait-until block has no jaws and nothing inside it. Is it still doing something the whole time? What?

Step 7 — Mission 13 Advanced — all three cubes13 pts

Use the touch sensor to make your approach repeatable. If your robot can find a wall or a cube by contact instead of by timing, every placement after that starts from a known spot.

Where in your run does the touch sensor help most — finding the cubes, or finding your way back?

• All three Unstraight Cubes are in one stack or pyramid
• It is still standing after the robot backs away

Step 8 — Run it five times

Did the stack fall over on any run? What knocked it — the robot leaving, or the stack itself?

Score It — Checkpoint

My score

With Project 6 and Project 8, Mission 13 is now worth 21 points across all three parts — the most any single mission has given you.

Finish the check

Say which blocks you would use for each.

Spot the bug

Each of these is wrong. Say what happens when you run it.

Can you do it again?

• I can explain what a hexagon block is and what its only two answers mean
• I read a sensor live in the app before I build any blocks with it
• I know what wait until does the whole time it is waiting
• I know what goes inside a repeat-until’s jaws and what goes after them
• I test a new sensor program with the robot in my hands first
• I can use contact with a wall to give my robot a position it can trust

Think about it

Since Project 3, every distance your robot has driven was really a length of time. What does a touch sensor give you that a stopwatch never could?

A touch sensor only tells you something when you are already touching it. Name one thing on the field you would want to stop at, where bumping into it first would be a bad idea.

Your loop asks a question over and over. So far the question has always been about a sensor. What else could a robot count or measure that a loop could watch?

Next

A wall tells you where you are, but only if there is a wall. Most of the field has nothing to bump into.

In Project 11 — Counting and Turning by Feel, you point the same checking at a number instead of a switch — the degrees your wheels have rolled, and the angle the whole robot has turned. The brick has been measuring both all along.