EASEBRAIN ACADEMY — OFFLINE MISSION PACK Sense. Test. Respond. https://easebrain.live/curriculum/sense-and-respond CHALLENGE Build or simulate a water-level indicator for a model tank. Define the threshold and repeat tests below, near and above it. LEARNING OBJECTIVES Distinguish measurement variation from a decision threshold. Record false alerts, missed alerts and missing readings. LEARN & APPLY 1. Know what you measure A water-level sensor estimates level or distance. It does not establish drinking safety or water quality. Label units and explain the relationship between a reading and your model. Practice: Draw input, processing and output with the measurement units. 2. Calibrate and repeat Compare readings with known positions. Repeat trials at low, near-threshold and high levels. Record inconsistent alerts as evidence rather than deleting them. Practice: Run five trials at each of three levels, or use an equivalent simulation. 3. Make failure visible Test a missing reading and a restart. Use only facilitator-approved low-voltage equipment; keep liquids away from electronics. A simulation is a complete route through this mission. Practice: Show a failure mode, one improvement and its retest. WORKED EXAMPLE A simulated tank alert turns on at 80 cm. Record the true model level, reading, expected alert and actual alert. A true level of 82 cm with no alert is a missed alert; a true level of 72 cm with an alert is a false alert. YOUR PRACTICE QUESTION Why might an alert flicker when a reading is close to 80 cm? Write your answer before reading the explanation below. EXPLANATION Small variations can repeatedly cross the boundary. A possible improvement is to turn on at 80 cm and off at 75 cm, then test whether this behaviour fits the purpose. Never hide missing readings as normal results. STRETCH Test two thresholds, a missing reading and restart in a simulator. No hardware purchase is needed. PRACTICAL: Stop an alarm from flickering Suggested time: 60–90 minutes Illustrated crew scenario (fictional learning example) My tank alert switches on and off near 80 cm. Keep a state. Turn on at 80, but do not turn off until the reading falls to 75. PREDICT: What happens at 78 cm if the alarm was already on? What if it was off? STARTER: Simulated water-height readings in cm: 74,79,80,78,76,75,74. Initial state off. Turn on at height ≥80; turn off at height ≤75. Treat missing readings as unknown. These classroom values are not an engineering safety specification. 1. Trace a simple threshold First label each reading on if ≥80 and off otherwise. Observe that it turns off immediately at 78. Explain why repeated values near 80 could cause flicker. 2. Add memory Use the sensor lab or a paper state table: retain the previous state between 75 and 80. Compare the two sequences. State is a stored value that can change future behaviour. 3. Test faults and restart Insert ‘?’ for missing data and explain why unknown must be visible. Restart the model at 78 and state your initial-state assumption. If using hardware later, seek supervised validation; a simulator is not a safety device. COMPARE AFTER TRYING: With memory, the sequence is off, off, on, on, on, off, off. At 78 the result depends on previous state. Missing data shows unknown; in this classroom model the last valid internal state is retained for the next reading. A real control system requires a separate safe-failure design. SELF-CHECK My state table matches all seven readings. I test a missing value and a restart, not just normal data. I explain sensor input, stored state and output separately. TRANSFER: Change thresholds to 70 and 85. Predict a new sequence before running it and explain the trade-off. WORKBOOK TEMPLATE Time | Reading (cm) | Previous state | Rule used | Output Simple threshold sequence: Two-threshold sequence: Missing-data behaviour: Restart assumption: EVIDENCE CRITERIA Clear diagram and units Repeated measurements and edge-case tests One justified revision and honest limitations Deliverable: Diagram, repeated tests and a maintenance note SESSION PLAN (adapt with your facilitator) Recall & plan: 30 minutes Model & discuss: 45 minutes Guided practice: 75 minutes Build & test: 120 minutes Feedback & revise: 60 minutes Reflect & next step: 30 minutes RESOURCES MakeCode micro:bit simulator — https://makecode.microbit.org/device/simulator micro:bit project library — https://www.microbit.org/projects/make-it-code-it/ AI DISCLOSURE TEMPLATE Tool/version; date; task; information shared; claims checked and sources; what I changed; what I can explain myself. No AI or equipment purchase is required. Keep personal information out of project evidence.