Two Sensors, One Decision

Overview

So far, each behavior used one sensor. Your line-follow watches the Tophat. Your touch-stop watched the button. Today the robot uses two sensors at the same time: the Tophat to stay on the line, and a new ET distance sensor to watch for an object ahead. The robot will follow the line — and the moment something appears in front of it, stop. Neither sensor could do this alone. Together, they can.

By the end of this activity you will be able to:

• Read an ET distance sensor with k.analog(1) and explain what its values mean.
• Map the ET’s value-to-distance relationship, and find its close-range blind spot.
• Combine two sensors in one loop — one to steer, one to decide when to stop.
• Connect multi-sensor decisions to how intelligent systems fuse many inputs.

Phase 1 — Meet the ET Sensor

Move an object from far to near while watching analog(1). Describe what the number did. Did it climb steadily, or faster at some distances than others?

Phase 2 — Map Value vs. Distance

Place an object squarely in front of the sensor at each distance and record the ET value (pick the middle of the bounce). This table is your map from “value” to “real distance.”

Now Investigate the Close Range

Carefully push the object in past 3 inches and watch what the value does. Record it — this is real data about where the sensor fails.

What happened to the value when you went closer than 3 inches? Why is it dangerous to trust the sensor in that range?

Phase 3 — Choose Your Stop Value

You want the robot to stop with the object close — but before it enters the untrustworthy blind spot. A safe target is around 4 inches: close enough to count as “reached the object,” but safely outside the bad zone under 3 inches.

Why did you pick a stop value from around 4 inches instead of 1 or 2 inches, even though closer would “reach” the object more?

Phase 4 — Concept: Two Sensors in One Loop

while k.analog(1) < STOP_VALUE:     # ET: still far? keep going
    if k.analog(0) > MIDPOINT:       # Tophat: steer on the line
        ...
    else:
        ...
    k.msleep(10)                     # tiny pause (like the touch-sensor lab)

Phase 5 — Build line_follow_until_object

Start from your tuned line-follow. Add an ET check to the loop condition and your stop value at the top. Use your own MIDPOINT, STOP_VALUE, and the mav speeds you found best. Define line_follow_until_object above main(), as always.

#!/usr/bin/python3

# Unit 2, Big Idea 6: Two Sensors, One Decision

# Name: _______________________   Date: ___________

import os, sys
sys.path.append("/usr/lib")
import _kipr as k

MIDPOINT   = ____   # your Tophat threshold (from BI4)
STOP_VALUE = ____   # your ET value at ~4 inches (from Phase 3)
FAST       = ____   # your best mav fast speed (from BI5)
SLOW       = ____   # your best mav slow speed (from BI5)

def main():
    line_follow_until_object()    # follow the line, stop at the object

def line_follow_until_object():
    while k.analog(1) < STOP_VALUE:    # ET: object still far? keep going
        if k.analog(0) > MIDPOINT:     # Tophat: on black, steer right
            k.mav(0, FAST)
            k.mav(1, SLOW)
        else:                          # on white, steer left
            k.mav(0, SLOW)
            k.mav(1, FAST)
        k.msleep(10)                   # tiny pause so we don't overwork the controller

    k.motor(0, 0)                      # The object is close, so brake.
    k.motor(3, 0)
    k.msleep(50)                       # let the brake settle

main()

Test Log

Checklist

• The loop condition reads the ET: k.analog(1) < STOP_VALUE
• The if/else inside reads the Tophat: k.analog(0) > MIDPOINT
• There is an k.msleep(10) inside the loop
• Your STOP_VALUE is a ~4 inch reading — outside the blind spot
• The robot brakes after the loop

Phase 6 — Connect: The AI Literacy Bridge

Your robot just did something it never could with one sensor: it stayed on a path and watched for an obstacle, at the same time. This is called sensor fusion, and it’s how every advanced intelligent system works. A self-driving car blends cameras, radar, and GPS at once — no single one is enough. Your phone blends the touchscreen, the accelerometer, and the light sensor to decide what to show. Intelligence grows when a system stops relying on one input and starts combining many.

Read each scenario. Think it through, then write your answer.

Your robot used the Tophat AND the ET together. Describe a task that needs both — something neither sensor could accomplish alone.

You learned the ET can’t be trusted closer than ~3 inches. Why is it important for an intelligent system to know not just what its sensors say, but when not to trust them?

A self-driving car combines cameras, radar, and GPS. Why is combining several sensors safer than relying on the single “best” one?

Phase 7 — Individual Reflection

Complete this section on your own.

1. How does the ET sensor’s value relate to distance? What’s surprising about it compared to “bigger number means farther”?

2. What is the ET’s blind spot, and how did it affect the stop value you chose?

3. In your loop, what question did the Tophat answer, and what question did the ET answer? Why did you need both?

4. Complete this in 2–3 sentences: “Intelligent systems combine multiple sources of information to make decisions. This means that the more a system can sense, the better it can…”

Extension Challenges

Finished early? Try one or more of these.

Extension A — Back Off to a Safe Distance

• After stopping, what if the object is a little too close? Add a short backward move so the robot ends at a consistent distance. How does this avoid the blind spot entirely?

Extension B — Report the Distance

• Use your Phase 2 table to turn the stopping ET value back into an approximate distance, and print it. Roughly how far away did the robot actually stop?

Extension C — Two Conditions to Keep Going

• What if you wanted the robot to also stop after a maximum number of ticks, even if it never sees an object? How could the loop check the ET and the encoder? (Think about combining conditions.)

Extension D — Add a Third Sensor

• Imagine adding the touch sensor from Big Idea 1 as a backup bumper. How would three sensors together make the robot even more reliable? Sketch the idea in words.

Extension E — Automatic Data Collection

• Your ET sensor automatically collects distance data the entire time your robot runs — no one approves each individual reading. Real devices do this constantly: traffic cameras, smart doorbells, fitness trackers.
• What privacy concerns come up when a device collects data automatically instead of only when someone asks it to? Who should get to see that data, and who should decide?

Extension F — A Simple Rule-Based Decision

• Right now your robot stops using one rule: distance < threshold. A basic rule-based decision system combines multiple pieces of evidence before acting, instead of reacting to a single reading.
• Add a second condition: only stop if the object reads close on three readings in a row, not just one (to ignore a single noisy blip). Does requiring repeated evidence reduce false stops?