562  Sensor Selection Scenarios

13 Real-World IoT Case Studies for Sensor Selection Practice

562.1 Learning Objectives

By studying these scenarios, you will be able to:

  1. Analyze real-world IoT application requirements systematically
  2. Apply sensor selection criteria to diverse use cases
  3. Justify sensor choices based on accuracy, cost, power, and reliability trade-offs
  4. Recognize common patterns in sensor selection across different industries

Each scenario presents a real-world IoT application with specific requirements. As you read:

  1. Identify the key constraints (accuracy, budget, power, environment)
  2. Think about which sensor you would choose before reading the solution
  3. Study the explanation to understand why the optimal sensor was selected
  4. Note the trade-offs - why other sensors were not chosen

These scenarios are also used in the Sensor Selection Challenge Game where you can practice making these decisions interactively.

562.2 Beginner Scenarios

These scenarios focus on common consumer and hobbyist applications with straightforward requirements.

562.2.1 Scenario 1: Smart Home Temperature Monitor

Application: WiFi-enabled temperature and humidity monitor for a living room. The device logs data every 5 minutes and displays it on a mobile app.

Requirement Specification
Measurement Temperature (-10 to 50C) and Humidity (20-80% RH)
Accuracy Plus or minus 1C and plus or minus 5% RH acceptable
Environment Indoor, stable conditions
Power USB powered (always on)
Budget Low cost priority ($20 per unit)
Lifespan 2+ years

Optimal Sensor: DHT22

Why: The DHT22 is optimal because:

  1. Integrated temp + humidity in one sensor saves cost and complexity
  2. Plus or minus 0.5C accuracy exceeds the plus or minus 1C requirement
  3. $5 cost fits tight budget easily
  4. 1-Wire interface is simple to integrate with WiFi MCU
  5. 2-3 year lifespan meets requirement
  6. USB power makes 1.5mA consumption negligible

Why Not SHT85? While more accurate, the $25 cost is 5x higher with no benefit for this non-critical home application. Over-specifying accuracy wastes budget.

562.2.2 Scenario 2: Automatic Room Lighting Control

Application: Office building wants automatic lights that turn on when people enter dark rooms. 200 sensors needed with battery backup power.

Requirement Specification
Measurement Ambient light (1-10000 lux range)
Accuracy Plus or minus 20% acceptable
Environment Indoor office, fluorescent lights
Power Low power critical (battery backup)
Budget Moderate ($500 total for 200 sensors)
Lifespan 10 years

Optimal Sensor: BH1750

Why: The BH1750 is optimal because:

  1. Wide 1-65535 lux range covers dark (1 lux) to bright office (1000 lux)
  2. $2 cost allows 200 sensors within $500 budget ($400 total)
  3. 120 microamp power consumption is critical for battery backup
  4. I2C interface is standard on building automation controllers
  5. 10-year lifespan avoids frequent maintenance

Why Not TSL2591? While more capable, the $8 cost (4x higher) exceeds budget, and the wide dynamic range is overkill for indoor lighting.

562.2.3 Scenario 3: Parking Spot Occupancy Detector

Application: Shopping mall needs 50 sensors to detect car presence in parking spots (2-3m range). Outdoor installation with temperature varying 0-40C.

Requirement Specification
Measurement Distance 0.5-3 meters
Accuracy Plus or minus 5 cm acceptable
Environment Outdoor parking, temp swings
Power Wired power available
Budget Low cost ($150 total)
Lifespan 3+ years

Optimal Sensor: HC-SR04

Why: The HC-SR04 is optimal because:

  1. 2-400cm range perfectly covers parking spot detection (2-3m)
  2. $2 cost allows 50 sensors within $150 budget ($100 total)
  3. Simple trigger/echo interface is easy to multiplex
  4. 15mA power is acceptable with wired supply
  5. Plus or minus 3mm accuracy far exceeds plus or minus 5cm requirement

Why Not VL53L1X? ToF laser is 7x more expensive ($15 vs $2) and provides excessive precision for simple car/no-car detection. Temperature compensation is doable with software.

562.2.4 Scenario 4: Fitness Tracker Motion Sensor

Application: Wearable fitness band needs to count steps, detect running vs. walking, and track sleep. Must be ultra-low power (coin cell battery, 1 year life).

Requirement Specification
Measurement Acceleration and rotation
Accuracy Plus or minus 5% step counting
Environment Body temperature, low vibration
Power Ultra-low power (<1mA average)
Budget Moderate ($10-15 per unit)
Lifespan Device lifetime (3 years)

Optimal Sensor: MPU-6050

Why: The MPU-6050 is optimal because:

  1. 6-axis (accel + gyro) enables step counting, activity classification, and sleep tracking with one sensor
  2. $5 cost fits moderate budget with margin for other components
  3. 3.9mA active power is acceptable when duty-cycled (sleep 90% of time gives <0.5mA average)
  4. Integrated DMP offloads processing from MCU, saving more power
  5. Small package size is critical for wearables

Why Not BNO055? The magnetometer adds cost ($30 vs $5) with no benefit for fitness tracking, and power is higher (12.3mA).

562.3 Intermediate Scenarios

These scenarios involve industrial applications with more complex trade-offs.

562.3.1 Scenario 5: Industrial HVAC CO2 Monitoring

Application: Factory HVAC system needs to maintain indoor CO2 below 1000 ppm for worker safety and productivity. 10 sensors across factory floor. Must be accurate for regulatory compliance.

Requirement Specification
Measurement CO2 concentration (400-2000 ppm)
Accuracy Plus or minus 50 ppm required
Environment Industrial, temp 15-35C
Power 24V DC available
Budget Good budget ($1000 for 10 sensors)
Lifespan 10 years, low maintenance

Optimal Sensor: SCD40

Why: The SCD40 is optimal because:

  1. True NDIR CO2 measurement (not estimated like CCS811)
  2. Plus or minus 40ppm plus or minus 5% accuracy meets plus or minus 50ppm regulatory requirement
  3. 10+ year lifespan with auto-calibration minimizes maintenance
  4. $50 cost allows 10 sensors within $1000 budget ($500 total)
  5. Factory calibration and long-term stability are critical for compliance

Why Not CCS811? eCO2 estimation from TVOC is inaccurate (spec says plus or minus 50ppm, but real error is often plus or minus 100ppm), and baseline drift requires frequent recalibration. MQ-135 is even worse (plus or minus 20-50% error).

562.3.2 Scenario 6: Weather Station Network

Application: City deploys 50 outdoor weather stations measuring temperature, humidity, and barometric pressure for weather forecasting. Must handle -20 to +50C, rain/snow exposure. Solar powered.

Requirement Specification
Measurement Temp, Humidity, Pressure
Accuracy Plus or minus 0.2C, plus or minus 2% RH, plus or minus 1 hPa
Environment Outdoor, extreme conditions
Power Solar (low power critical)
Budget Moderate ($75 per station)
Lifespan 10 years outdoor

Optimal Sensor: SHT85 + BMP388

Why: The SHT85 + BMP388 combination is optimal because:

  1. SHT85 provides plus or minus 0.1C temp and plus or minus 1.5% humidity (exceeds spec)
  2. BMP388 provides plus or minus 0.33 hPa pressure (exceeds plus or minus 1 hPa requirement)
  3. Combined cost $25 + $8 = $33, well under $75 budget
  4. Both are ultra-low power (600 microamps + 3.4 microamps = <1mA average) critical for solar
  5. Both rated -40 to +85C and 10+ year lifespan for outdoor deployment

Why Not DHT22? Fails outdoor requirements: poor accuracy (plus or minus 0.5C, plus or minus 2%), only 2-year lifespan, and -40C operation is not guaranteed.

562.3.3 Scenario 7: Drone Obstacle Avoidance

Application: Racing drone needs forward-facing distance sensor to avoid trees/walls at 20 mph (9 m/s). Must measure 0.5-4m range, update 20Hz+, and work in bright sunlight and darkness.

Requirement Specification
Measurement Distance 0.5-4 meters
Accuracy Plus or minus 5 cm
Environment Outdoor, all lighting conditions
Power Battery powered
Budget Moderate ($30 per drone)
Response Time <50ms (20Hz update)

Optimal Sensor: VL53L1X

Why: The VL53L1X is optimal because:

  1. ToF laser works in full sunlight (unlike ultrasonic which fails outdoors)
  2. 40-4000mm range perfectly covers obstacle detection at speed
  3. 20-50ms response time achieves 20Hz update rate
  4. Immune to ambient light (critical for racing drones)
  5. $15 cost fits budget
  6. 19mA power is acceptable for 5-10 minute drone flights
  7. Lightweight sensor (<2g)

Why Not HC-SR04? Ultrasonic fails outdoors (wind interference), is too slow (trigger-based, not continuous), and is temperature-sensitive.

562.3.4 Scenario 8: Robot Arm Orientation Tracking

Application: 6-axis robot arm needs precise orientation feedback for each joint (6 sensors total). Industrial application requires plus or minus 1 degree accuracy, 100Hz update rate, and immunity to magnetic interference from motors.

Requirement Specification
Measurement 3D orientation (Euler angles)
Accuracy Plus or minus 1 degree absolute heading
Environment Factory, strong magnetic fields
Power Wired power available
Budget Good ($300 for 6 sensors)
Response Time 100 Hz fusion output

Optimal Sensor: BNO055

Why: The BNO055 is optimal because:

  1. 9-axis fusion provides absolute orientation (not just relative motion like MPU-6050)
  2. Plus or minus 1 degree heading accuracy meets spec exactly
  3. 100Hz fusion output matches required update rate
  4. Integrated ARM processor offloads sensor fusion from robot controller
  5. $30 cost allows 6 sensors within $300 budget ($180 total)
  6. Auto-calibration is critical for industrial reliability

Why Not MPU-6050? Only provides relative motion (gyro integrates to drift), no absolute heading, and 6-axis cannot determine yaw without magnetometer.

562.4 Advanced Scenarios

These scenarios represent challenging industrial and research applications with strict requirements.

562.4.1 Scenario 9: Semiconductor Cleanroom Monitoring

Application: ISO Class 3 cleanroom for chip fabrication requires ultra-precise temp (plus or minus 0.05C) and humidity (plus or minus 1% RH) control. Even 0.1C drift causes wafer defects ($100k+ loss). 20 sensors, critical reliability.

Requirement Specification
Measurement Temperature and Humidity
Accuracy Plus or minus 0.05C, plus or minus 1% RH (critical)
Environment Cleanroom, 20-22C stable
Power Wired power
Budget High ($1500 for 20 sensors)
Lifespan 10 years, zero drift

Optimal Sensor: SHT85

Why: The SHT85 is optimal (even mandatory) because:

  1. Plus or minus 0.1C accuracy is the ONLY sensor that meets plus or minus 0.05C after calibration
  2. Plus or minus 1.5% RH achieves plus or minus 1% spec
  3. Factory-calibrated with 10-year stability ensures zero drift (critical for $100k wafer batches)
  4. $25 cost allows 20 sensors within $1500 budget ($500 total)
  5. I2C allows networked monitoring with central control

Why Not DHT22? Completely inadequate: plus or minus 0.5C is 10x worse than spec, and 2-year lifespan with drift is unacceptable. This is the rare case where only one sensor works.

562.4.2 Scenario 10: Solar Panel Sun Tracking System

Application: Utility-scale solar farm needs light sensors to track sun position across 0.01-100,000 lux range (pre-dawn to midday). Must measure both visible and IR. 100 dual-axis trackers.

Requirement Specification
Measurement Light intensity (wide dynamic range)
Accuracy Plus or minus 10% across 0.01-100k lux
Environment Outdoor, extreme temp/sun
Power Solar powered (low power)
Budget Moderate ($2000 for 100 sensors)
Lifespan 15 years outdoor

Optimal Sensor: TSL2591

Why: The TSL2591 is optimal because:

  1. Ultra-wide 0.000188-88000 lux range captures pre-dawn to direct sun (BH1750 saturates >65k lux)
  2. Separate IR and visible channels enable true sun tracking (compensate for clouds)
  3. $8 cost allows 100 sensors within $2000 budget ($800 total)
  4. 400 microamp power is acceptable with solar
  5. 10+ year lifespan for outdoor deployment

Why Not BH1750? Saturates in direct sunlight (>65k lux), cannot distinguish IR, and provides less accurate tracking (plus or minus 20% vs plus or minus 10%). For solar farms, TSL2591’s extra cost is justified by 5-10% energy gain.

562.4.3 Scenario 11: Autonomous Warehouse Robot Navigation

Application: 100 warehouse robots need precise 3D orientation for SLAM (simultaneous localization and mapping). Plus or minus 0.5 degree heading accuracy required. Magnetic interference from metal shelves and forklifts.

Requirement Specification
Measurement 9-axis orientation (SLAM)
Accuracy Plus or minus 0.5 degree heading
Environment Indoor, strong magnetic interference
Power Battery powered (8 hour shifts)
Budget High ($5000 for 100 robots)
Response Time 50 Hz minimum

Optimal Sensor: BNO055

Why: The BNO055 is optimal because:

  1. 9-axis fusion with magnetometer provides absolute heading for SLAM (MPU-6050 drifts in minutes)
  2. Plus or minus 1 degree heading accuracy is achievable with calibration (close to plus or minus 0.5 degree spec)
  3. 100Hz fusion output exceeds 50Hz requirement with margin
  4. $30 cost allows 100 sensors within $5000 budget ($3000 total)
  5. Auto-calibration compensates for warehouse magnetic interference
  6. 12.3mA power is acceptable for 8-hour battery operation with sleep modes

Why Not MPU-6050? Gyro drift makes absolute heading impossible for long-duration SLAM - robots would lose position within 10 minutes.

562.4.4 Scenario 12: Research-Grade Air Quality Station

Application: University atmospheric research station needs ppb-level gas detection (CO2, VOCs) for climate studies. Data published in peer-reviewed papers. Outdoor -40 to +50C. 10-year deployment.

Requirement Specification
Measurement CO2 (plus or minus 10 ppm) and VOC
Accuracy Research-grade (plus or minus 1%)
Environment Outdoor research station
Power Solar + battery backup
Budget Research grant ($2000 per station)
Lifespan 10 years, minimal maintenance

Optimal Sensor: SCD40 (+ CCS811 for TVOC)

Why: The SCD40 is optimal because:

  1. True NDIR CO2 with plus or minus 40ppm accuracy far exceeds research-grade needs
  2. 10+ year lifespan with auto-calibration minimizes maintenance at remote station
  3. $50 cost is negligible with research grant budget
  4. Factory calibration and long-term stability are critical for publishable data
  5. 18mA power is acceptable with solar + battery backup
  6. Pair with CCS811 for TVOC to get both CO2 and VOC (total $62)

Why Not MQ-135? Completely inadequate for research: plus or minus 20-50% error, cross-sensitivity, frequent calibration, and short lifespan make data unusable for publications.

562.4.5 Scenario 13: High-Speed Production Line Inspection

Application: Automotive parts move at 2 m/s on conveyor belt. ToF sensor must measure part height (5-50mm range) with plus or minus 0.5mm accuracy to detect defects. 100 measurements/second. Industrial environment with vibration and dust.

Requirement Specification
Measurement Distance 5-50 mm
Accuracy Plus or minus 0.5 mm (high precision)
Environment Industrial, vibration, dust
Power 24V industrial power
Budget Industrial budget ($100 per sensor)
Response Time 100 Hz (10 ms)

Optimal Sensor: VL53L1X

Why: The VL53L1X is optimal because:

  1. ToF laser achieves plus or minus 25mm @ 3m accuracy, which scales to <plus or minus 0.5mm at 50mm range
  2. 20-50ms response time allows 100Hz measurement rate for fast-moving parts
  3. Laser-based measurement is immune to ambient light and vibration (unlike ultrasonic)
  4. $15 cost fits industrial budget well
  5. I2C interface is standard in industrial PLCs
  6. No moving parts for reliability in dusty environments

Why Not HC-SR04? Ultrasonic is too slow (trigger-based), has insufficient accuracy, and is sensitive to vibration/temperature in factory. This application demands laser precision.

562.5 Summary

These 13 scenarios demonstrate key principles of sensor selection:

Principle Example Scenario
Match accuracy to requirements Home thermostat uses DHT22, not SHT85
Calculate 5-year TCO MQ-135 costs $406 vs SCD40’s $50 over 5 years
Consider power constraints BH1750’s 120uA enables battery backup
Check environmental ratings SHT85 survives outdoor; DHT22 does not
Verify response time VL53L1X achieves 100Hz; HC-SR04 cannot
Account for interference BNO055’s fusion handles magnetic noise

562.6 What’s Next

Practice applying these principles interactively:

Sammy the Sensor explains: “Picking the right sensor is like picking the right tool. You wouldn’t use a sledgehammer to hang a picture! Use the simplest sensor that meets your requirements - it saves money and usually works better.”

Lila the Light Sensor adds: “I learned that sometimes cheap sensors fail in tough conditions. For outdoor weather stations, spend a bit more on industrial-grade sensors. They last 10 years instead of 2!”