Robot Logic in Scratch

Lessons

Senior 3 · ICT

Robot Logic in Scratch

5 lessons

Lesson 5 of 5

Combining Sensors: Conditions, Nested Ifs and Sensor Fusion

60 min

One sense is never enough

A robot vacuum cleaner carries bumpers to feel furniture, sensors underneath to notice the top of a staircase, sensors in its wheels to count how far it has gone, and often more. No single sensor could keep it safe: the bumper cannot see a staircase, and the stair sensor cannot feel a chair. The robot's program has to combine all of them, decide which matters most, and sometimes merge several readings into one better estimate. This lesson does both: combined conditions with if, if-else and nested if, and sensor fusion.

Definition

Compound condition: two or more conditions joined with and, or or not, such as “left beam sees a wall AND right beam sees a wall”.

Nested if: an if block placed inside another if (or if-else) block, so a second question is asked only after the first has been answered.

Priority: the order in which conditions are checked; the most urgent goes first.

Sensor fusion: combining the readings of several sensors to get information that is more accurate, more reliable or more complete than any one sensor gives alone.

Three ways to decide

Block	Meaning	Robot example
if … then	Do something only when the condition is true; otherwise do nothing extra.	If the loudness is over 50, say “Heard you!”.
if … then … else	Choose between exactly two actions.	If the beam sees a wall, turn; else move forward.
nested if	Ask further questions inside one branch, giving three or more outcomes.	If the bumper is pressed, reverse; else, if both beams see walls, turn round; else …

Joining conditions: and, or, not

The green and, or and not blocks in Operators combine true-or-false conditions. Their results:

A	B	A and B	A or B	not A
false	false	false	false	true
false	true	false	true	true
true	false	false	true	false
true	true	true	true	false

and is true only when both parts are true; or is true when at least one part is true; not flips true and false.

A robot with three sensors

Give the robot costume two short beams, a cyan one angled to the front-left and a pink one to the front-right, and keep the bumper (touching the Wall sprite). The script stores each beam's reading in a variable, then decides in order of priority:

Block diagram drawn by Ganzaa. In Scratch, a true-or-false block can be dropped into the set block; the variable then holds the word true or false.

Bumper	Left beam	Right beam	Branch	Action
pressed	any	any	①	Reverse 10 steps, turn round
not pressed	wall	wall	②	Turn round: dead end
not pressed	wall	clear	③	Turn right 15°, away from the left wall
not pressed	clear	wall	④	Turn left 15°
not pressed	clear	clear	⑤	Move forward 3 steps

The bumper is checked first because touching means the other sensors have already failed: it is the most urgent. Two sensors give 4 combinations; with the bumper, the nested ifs sort all 8 possible states into 5 actions.

Worked example 1: tracing the decisions

Moment	Bumper	left	right	Tests in order	Action
A	no	false	false	① no; ② no; ③ no; ④ no	⑤ move forward
B	no	true	false	① no; ② no (right is false); ③ yes	turn right 15°
C	no	true	true	① no; ② yes	turn round
D	yes	true	true	① yes: nothing else is checked	reverse and turn round

Notice moment D: both beams also see walls, but the bumper wins because it is tested first. If ② were tested before ①, the robot would turn on the spot while pressed against the wall, scraping it. Order is part of the logic.

Sensor fusion

Every sensor has weaknesses. An ultrasonic sensor can miss soft curtains, which absorb sound, or a wall met at a steep angle, which reflects the echo away. An infrared distance sensor can be confused by bright sunlight or very dark surfaces. A camera sees little in the dark. Fusing sensors whose weaknesses differ gives a better picture.

Method	How it works	Example
Average	Add the readings and divide by how many there are	Two distance sensors pointing the same way
Weighted average	Trust the more accurate sensor more	0.75 × ultrasonic + 0.25 × infrared
Voting or median	With three or more sensors, take the middle value, so one faulty sensor cannot spoil the result	Three thermometers in a greenhouse
Agreement (and)	Act only when two different sensors agree	Fire alarm: heat and smoke
Coverage (or)	Act when any sensor detects something	Stop if the bumper or the stair sensor triggers

Fusion is everywhere: a phone combines its accelerometer, gyroscope and compass to know which way it points; self-driving cars combine cameras, radar and lidar.

Block diagrams drawn by Ganzaa. In a simulation, sonar, infrared, temperature and smoke are slider variables.

Worked example 2: fusing readings

Weighted average. The ultrasonic sensor reads 52 cm and the infrared sensor 60 cm. The ultrasonic one is known to be more accurate, so it gets weight 0.75 and the infrared 0.25 (the weights add up to 1):

$d = 0.75 \times 52 + 0.25 \times 60 = 39 + 15 = 54 \text{ cm}$

Median. Three thermometers read 31 °C, 32 °C and 75 °C; the third has a fault. The mean is $\frac{31 + 32 + 75}{3} = 46$ °C, which would wrongly switch on every fan. The median (the middle value when sorted) is 32 °C, which is right. With three sensors, the median ignores one bad reading.

Agreement. A kitchen gets hot when cooking (55 °C) but there is no smoke (10): temperature > 50 and smoke > 30 is false, so no false alarm. With or, the alarm would ring at every meal.

Warning

Common logic mistakes

and when you meant or: “stop if the bumper and the stair sensor trigger” would drive off a staircase when the bumper is not pressed. Safety stops normally use or.
Wrong order: put the most urgent, safety-critical test first.
Overlapping ifs instead of else: two separate if blocks can both run in one frame (turn right and then turn left). Use if-else so exactly one action is chosen.
Equal weights for unequal sensors: a plain average trusts a poor sensor as much as a good one.

MATLAB: Understanding Sensor Fusion and Tracking, Part 1: What Is Sensor Fusion? An engineer explains why autonomous systems combine sensors, with clear examples. Watch at least the first half.

✦

Hands-on: a smarter robot

In Robot_Sim, give the robot costume two short beams (cyan front-left, pink front-right) and keep a Wall sprite for the bumper.
Build the nested-if script. Show the left and right monitors and watch them change.
Fill in the decision table above from your own tests: for each of the five branches, make a situation that triggers it.
Swap the order of tests ① and ② and describe what goes wrong.
Fusion: make slider variables sonar and infrared (0–200) and the fusion script. Set sonar to 30 and infrared to 90: what does fused show, and does the robot warn? Try changing the weights.
Fire alarm: make sliders temperature and smoke, and a sprite that shows “FIRE!” when it receives fire alarm. Test four cases: cool and clear, hot and clear, cool and smoky, hot and smoky. Then change and to or and test again. Which is better, and why?
Present your robot to the class: show one branch of each kind and explain one fusion rule.

Practice

Define a nested if, a compound condition and sensor fusion.
Complete: A = true, B = false. Give A and B; A or B; not B; (not A) or B.
Explain the difference between if-then and if-then-else, with a robot example of each.
In the three-sensor robot, which action happens when the bumper is not pressed, left is false and right is true? Which tests are checked?
An ultrasonic sensor reads 80 cm and an infrared one 70 cm. With weights 0.6 and 0.4, find the fused distance.
Five greenhouse thermometers read 24, 25, 25, 26 and 61 °C. Find the mean and the median, and say which the controller should use.
A pupil's robot sometimes turns right and then immediately left in the same moment. Analyse the bug in her script.
A school wants a robot that patrols the laboratory and calls for help in a fire. Design and justify a fusion rule using a temperature sensor, a smoke sensor and a camera, and evaluate one weakness of your rule.

Show answersHide answers

Nested if: an if inside another if or if-else, so a second question is asked only in one branch. Compound condition: conditions joined by and, or or not. Sensor fusion: combining several sensors' readings to get better information than any one alone.
A and B = false; A or B = true; not B = true; (not A) or B = false or false = false.
if-then does something only when the condition is true (if loudness > 50, say “Heard you!”); if-then-else chooses between two actions (if the beam sees a wall, turn; else move forward).
Branch ④: turn left 15°. Tests: ① bumper (no), ② both beams (no), ③ left only (no), ④ right (yes).
$0.6 \times 80 + 0.4 \times 70 = 48 + 28 = 76$ cm.
Mean: $\frac{24 + 25 + 25 + 26 + 61}{5} = \frac{161}{5} = 32.2$ °C. Median: 25 °C. Use the median: the 61 °C reading is clearly faulty and drags the mean up.
She used two separate if blocks (if left then turn right; if right then turn left) instead of if-else. When both beams see walls, both ifs run in the same frame. Use nested if-else so exactly one action is chosen, and handle “both” first.
Example rule: alarm if (temperature > 55 and smoke > 30) or (the camera sees flames and smoke > 30). Requiring two kinds of evidence avoids false alarms from a hot Bunsen burner or steam, while either pair still catches a real fire. Weakness: a smouldering fire with lots of smoke but little heat and no visible flame might be detected late; a safer design could also alarm on very high smoke alone.

1One sense is never enough

2Three ways to decide

3Joining conditions: and, or, not

4A robot with three sensors

5Sensor fusion

6Hands-on: a smarter robot