6  Sensor Hardware Selection

6.1 Learning Objectives

• Select sensors and microcontrollers matched to application domains (smart cities, agriculture, healthcare, industrial)
• Calculate battery life using average current consumption with active/sleep duty cycles
• Design Bills of Materials (BOMs) with complete cost breakdowns including sensors, MCU, connectivity, power, and enclosures
• Choose wireless connectivity protocols (Zigbee, Wi-Fi, LoRaWAN, NB-IoT) based on range, data rate, and power budget
• Apply cost optimization strategies across prototyping, small production, and mass production scales

In 60 Seconds

Choosing the right sensors and hardware for an IoT deployment requires matching components to your specific domain, environment, power budget, and cost constraints. This guide provides practical recommendations across smart cities, agriculture, home automation, healthcare, and industrial domains – complete with Bill of Materials (BOMs), microcontroller selection matrices, connectivity options, and power budget calculations. A smart parking node costs approximately $93, a soil moisture station $59, and a basic 8-sensor home system just $30.

Key Concepts

• Sensor Module vs. Bare Sensor: Sensor modules integrate the sensing element, signal conditioning, ADC, and communication interface on a small PCB; bare sensors require external conditioning circuits; modules cost more but dramatically reduce design time and risk
• Breakout Board: A PCB adapting a small SMD sensor package to breadboard-compatible through-hole pins with required passive components (decoupling caps, pull-up resistors) already populated; ideal for prototyping
• Development Board Ecosystem: Microcontroller development boards (ESP32 DevKit, Arduino Uno, Raspberry Pi Pico) provide USB programming, regulated power, and header pins; pre-built sensor shield expansions further reduce integration effort
• Shield / HAT: A daughter board that stacks on a development board (Arduino shield, Raspberry Pi HAT) providing ready-made sensor, display, or actuator functionality through a standardized mechanical and electrical interface
• Sensor Fusion Hardware: Multi-sensor ICs combining accelerometer, gyroscope, and magnetometer (9-axis IMU) or environmental parameters (BME688: temperature, humidity, pressure, gas) on a single chip, reducing BOM count and PCB area
• Datalogger Hardware: Systems combining sensors with local storage (SD card, flash) and a real-time clock for standalone deployment; essential when continuous connectivity is not available or when local data backup is required
• Sensor Node Design: A complete IoT sensing unit: sensor(s) + microcontroller + wireless module + power management + enclosure; designed for deployment in the target environment with appropriate IP rating, battery life, and connectivity
• Hot-Swap and Field-Replaceable Design: Hardware designed to allow sensor replacement without specialized tools or soldering; important for field maintenance of deployed sensor networks, especially in hard-to-access locations

For Beginners: Sensor Hardware Selection

Choosing hardware for an IoT project is like assembling a toolkit for a specific job. You need to match your sensors, microcontroller, wireless module, and power source to your application. A sensor for monitoring soil moisture on a farm has very different requirements (long battery life, weatherproof, long-range wireless) than one for a smart home (always powered, short-range Wi-Fi). This guide helps you match components to your specific use case.

6.2 Hardware Selection Guide by Application Domain

Time: 20 min | Difficulty: Advanced | Unit: P06.C03.U16

Selecting the right sensors and hardware for your specific IoT application domain requires understanding the unique requirements, constraints, and priorities of each use case. This guide provides practical recommendations for sensor selection across different application domains.

Application-Specific Hardware Selection

6.2.1 Smart Cities Hardware Recommendations

Application	Primary Sensors	Microcontroller	Connectivity	Power	Estimated Cost/Node
Smart Parking	Magnetic field sensor (PNI RM3100)	ESP32	LoRaWAN/NB-IoT	Battery (5-10 yr)	$80-120
Traffic Monitoring	Magnetic field + Camera	Raspberry Pi 4	Wi-Fi/4G	Mains	$150-300
Smart Lighting	LDR/BH1750 + PIR motion	ESP32	Wi-Fi/Zigbee	Mains	$25-40
Waste Management	HC-SR04 ultrasonic	ESP32	LoRaWAN	Solar + battery	$60-90
Air Quality	MQ-135, MH-Z19 (CO2), PMS5003	ESP32	Wi-Fi/4G	Solar + battery	$80-150
Noise Monitoring	MEMS microphone (SPH0645)	ESP32	Wi-Fi/4G	Solar + battery	$50-80

Key Considerations:

• Coverage: Smart parking needs 1 sensor per space; air quality 1 per km² in urban areas
• Battery Life: Parking sensors need 5-10 years; waste management 3-5 years
• Connectivity: LoRaWAN for wide area (10 km range); Wi-Fi for dense urban
• Cost: City-wide deployments require cost under $100/node for viability

Example BOM: Smart Parking Sensor Node

• Magnetic sensor PNI RM3100: $25
• ESP32 (low-power variant): $5
• LoRaWAN module (RFM95W): $8
• Lithium battery 3.6V 19Ah: $30
• Enclosure IP67: $15
• Installation hardware: $10
• Total: ~$93 per parking space

---

6.2.2 Smart Agriculture Hardware Recommendations

Application	Primary Sensors	Microcontroller	Connectivity	Power	Estimated Cost/Node
Soil Moisture	Capacitive sensor (STEMMA)	ESP32	LoRaWAN	Solar + battery	$40-60
Weather Station	BME280, anemometer, rain gauge	ESP32	Wi-Fi/4G	Solar + battery	$100-200
Irrigation Control	Soil moisture + flow sensor	ESP32	Wi-Fi	Solar + battery	$80-120
Greenhouse Monitoring	DHT22, CO2 (MH-Z19), light	ESP32	Wi-Fi	Mains	$60-100
Livestock Tracking	GPS + accelerometer	ESP32	LoRaWAN/GSM	Battery (6 mo)	$50-80

Key Considerations:

• Outdoor Rating: IP65+ enclosures required for field deployment
• Power: Solar panels (5-10W) with 10-20 Ah batteries for year-round operation
• Wireless Range: Farms need 1-5 km range (LoRaWAN ideal)
• Durability: Sensors exposed to UV, rain, temperature extremes (-20 to 50°C)

Example BOM: Soil Moisture Monitoring Station

• 4× Capacitive soil moisture sensors: $20
• ESP32 Dev Board: $5
• LoRaWAN RFM95: $8
• Solar panel 6V 3.5W: $12
• 18650 Li-ion battery 3000mAh: $6
• Waterproof box IP65: $8
• Total: ~$59 per monitoring station

---

6.2.3 Smart Home Hardware Recommendations

Application	Primary Sensors	Microcontroller	Connectivity	Power	Estimated Cost/Node
Security System	PIR + door/window sensors	ESP32	Wi-Fi	Battery (2-3 yr)	$30-50
Energy Monitoring	CT clamps (SCT-013)	ESP32	Wi-Fi	Mains	$40-60
HVAC Automation	DHT22 + occupancy (PIR)	ESP32	Wi-Fi/Zigbee	Mains	$25-40
Water Leak Detection	Water detection sensor	ESP8266	Wi-Fi	Battery (1-2 yr)	$15-25
Indoor Air Quality	BME680 + MH-Z19 (CO2)	ESP32	Wi-Fi	Mains/USB	$50-80

Key Considerations:

• Wi-Fi Coverage: Ensure 2.4 GHz Wi-Fi reaches all sensors (range extenders may be needed)
• Battery vs. Mains: Battery sensors for doors/windows; mains for energy monitors
• Integration: Choose ESP32/Zigbee for Home Assistant, MQTT compatibility
• User-Friendly: Simple setup, OTA updates, mobile app control

Example BOM: Complete Smart Home Sensor Kit

• 1× ESP32 hub: $5
• 3× PIR motion sensors: $6
• 2× Door/window sensors: $4
• 1× DHT22 temp/humidity: $5
• 1× MQ-2 gas sensor: $3
• 1× Water leak sensor: $2
• Enclosures, wiring: $5
• Total: ~$30 for basic 8-sensor home system

---

6.2.4 Smart Health Hardware Recommendations

Application	Primary Sensors	Microcontroller	Connectivity	Power	Estimated Cost/Node
Heart Rate Monitor	MAX30102 (pulse oximeter)	nRF52840	BLE	Battery (7-14 days)	$30-50
Fall Detection	MPU6050 (accelerometer/gyro)	ESP32	Wi-Fi/BLE	Battery (6-12 mo)	$25-40
Body Temperature	MLX90614 (IR non-contact)	ESP32	Wi-Fi/BLE	Battery/USB	$35-55
Medication Adherence	Load cell + RFID	ESP32	Wi-Fi	Mains/USB	$40-60
Sleep Monitoring	Pressure mat + MPU6050	ESP32	Wi-Fi	Mains	$50-80

Key Considerations:

• Medical Grade vs. Wellness: FDA approval needed for medical claims (costly); wellness ok for consumer
• Privacy: HIPAA compliance if storing health data (encryption, secure transmission)
• Battery Life: Wearables need 7+ days; bedside monitors can use mains power
• Accuracy: Heart rate ±2 bpm; temperature ±0.2°C for medical use

Example BOM: Wearable Heart Rate + Activity Monitor

• MAX30102 pulse oximeter: $8
• MPU6050 accelerometer/gyro: $4
• nRF52840 BLE module: $12
• 3.7V 250mAh LiPo battery: $5
• TP4056 charging module: $2
• Custom 3D printed case: $3
• Total: ~$34 for wearable device

---

6.2.5 Smart Industrial Hardware Recommendations

Application	Primary Sensors	Microcontroller	Connectivity	Power	Estimated Cost/Node
Vibration Monitoring	ADXL345 (3-axis accelerometer)	STM32F103	Modbus/Ethernet	Mains/24V DC	$80-150
Temperature Monitoring	Multiple DS18B20 (1-Wire)	ESP32	Wi-Fi/Ethernet	Mains	$50-80
Current Monitoring	SCT-013 CT clamps	ESP32	Wi-Fi/Modbus	Mains	$60-100
Machine Vision	Camera (ESP32-CAM)	ESP32	Wi-Fi	Mains	$40-80
Gas Detection	MQ-4 (methane), MQ-7 (CO)	ESP32	Wi-Fi/4G	Mains	$50-90

Key Considerations:

• Industrial Protocols: Support Modbus RTU/TCP, OPC UA, MQTT for integration
• Harsh Environments: IP65-67 rated enclosures, -40 to +85°C operating range
• Reliability: Industrial-grade components, redundancy, UPS backup
• Real-time: Sub-second response for safety-critical applications

Example BOM: Predictive Maintenance Vibration Monitor

• ADXL345 accelerometer: $8
• STM32F103 microcontroller: $3
• RS485 to TTL module: $5
• 24V to 5V DC-DC converter: $6
• Industrial DIN rail enclosure: $25
• Mounting bracket: $8
• Total: ~$55 per machine monitor

---

6.2.6 Smart Environment Hardware Recommendations

Application	Primary Sensors	Microcontroller	Connectivity	Power	Estimated Cost/Node
Air Quality	PMS5003, MH-Z19, BME680	ESP32	Wi-Fi/LoRaWAN	Solar + battery	$100-180
Water Quality	pH, DO, turbidity probes	ESP32	4G/LoRaWAN	Solar + battery	$300-600
Seismic Monitoring	ADXL355 (high-g accelerometer)	Raspberry Pi	4G	Mains/battery	$150-300
Forest Fire Detection	MQ-2, DHT22, smoke detector	ESP32	LoRaWAN	Solar + battery	$60-100
Noise Pollution	SPH0645 MEMS mic	ESP32	Wi-Fi/4G	Solar + battery	$50-90

Key Considerations:

• Remote Locations: Solar power mandatory; 4G/satellite for connectivity
• Calibration: Water quality sensors need monthly calibration
• Weather Resistance: IP67+ rating, conformal coating on PCBs
• Data Frequency: Seismic 100+ Hz; air quality 0.1 Hz

Example BOM: Environmental Air Quality Station

• PMS5003 PM2.5/PM10: $25
• MH-Z19C CO2 sensor: $20
• BME680 (temp/hum/VOC): $15
• ESP32: $5
• LoRaWAN RFM95: $8
• Solar panel 10W + controller: $25
• 12V 7Ah sealed lead-acid battery: $18
• Weatherproof enclosure: $30
• Total: ~$146 per monitoring station

---

6.2.7 Microcontroller Selection by Domain

Domain	Recommended MCU	Why?	Alternative
Smart Cities	ESP32 + LoRaWAN	Long range, Wi-Fi fallback, low power	nRF52840 (BLE mesh)
Agriculture	ESP32	Wi-Fi + BLE, 18 ADC channels for sensors	STM32 (industrial)
Home Automation	ESP8266/ESP32	Low cost, Wi-Fi, huge community	Zigbee modules
Healthcare	nRF52840	Ultra-low power BLE, wearable-friendly	ESP32 (if Wi-Fi needed)
Industrial	STM32F103/F4	Real-time, industrial protocols, rugged	ESP32 (with RS485)
Environment	ESP32	Multi-sensor, Wi-Fi/LoRaWAN options	Raspberry Pi (edge AI)

---

6.2.8 Connectivity Selection Matrix

Range Needed	Data Rate	Power Budget	Recommended Protocol	Hardware Module
< 100m	Low-Med	Very Low	Zigbee	CC2530, XBee
< 100m	Med-High	Low	Wi-Fi 2.4GHz	ESP32, ESP8266
< 100m	Very High	Medium	Wi-Fi 5GHz	Raspberry Pi 4
1-10 km	Very Low	Ultra Low	LoRaWAN	RFM95W, SX1276
1-10 km	Low	Low	NB-IoT	SIM7020, BC95
Unlimited	Med-High	Medium-High	4G LTE	SIM7600, EC25

Knowledge Check: Connectivity Selection

---

6.2.9 Power Budget Planning

Battery Life Estimation Formula:

Battery Life (hours) = Battery Capacity (mAh) / Average Current (mA)

Example: 2500 mAh battery, ESP32 sensor node
- Active (1 sec): 80 mA
- Sleep (59 sec): 10 µA = 0.01 mA
- Average = (80×1 + 0.01×59) / 60 = 1.34 mA
- Life = 2500 / 1.34 = 1865 hours = 78 days

With solar (10W panel):
- Daily generation: 10W × 4 hours (effective sunlight) = 40 Wh
- Daily consumption: 1.34 mA × 3.7V × 24h = 0.12 Wh
- Result: Solar provides 333× needed power → indefinite operation

Power Budgeting Tips:

1. Deep Sleep is Critical: ESP32 active (80 mA) vs. deep sleep (10 µA) = 8000× difference
2. Sensor Selection: Digital sensors (I2C) use less power than analog sensors requiring continuous ADC
3. Transmission Cost: Wi-Fi transmission (170 mA) vs. LoRaWAN (40 mA) = 4× difference
4. Solar Sizing: Minimum 5W panel for ESP32 with daily transmission in temperate climates

viewof batteryCapacity = Inputs.range([500, 10000], {
  value: 2500,
  step: 100,
  label: "Battery Capacity (mAh):"
})

viewof activeCurrent = Inputs.range([10, 200], {
  value: 80,
  step: 5,
  label: "Active Current (mA):"
})

viewof activeTime = Inputs.range([0.1, 60], {
  value: 1,
  step: 0.1,
  label: "Active Time per Cycle (seconds):"
})

viewof sleepCurrent = Inputs.range([0.001, 1], {
  value: 0.01,
  step: 0.001,
  label: "Sleep Current (mA):"
})

viewof cycleTime = Inputs.range([10, 3600], {
  value: 60,
  step: 10,
  label: "Cycle Time (seconds):"
})

sleepTime = Math.max(0, cycleTime - activeTime)

averageCurrent = (activeCurrent * activeTime + sleepCurrent * sleepTime) / (cycleTime > 0 ? cycleTime : 1)

batteryLifeHours = batteryCapacity > 0 ? batteryCapacity / (averageCurrent > 0 ? averageCurrent : 0.001) : 0

batteryLifeDays = batteryLifeHours / 24

md`### Battery Life Calculator

**Average Current:** ${averageCurrent.toFixed(2)} mA
**Battery Life:** ${batteryLifeHours.toFixed(0)} hours (${batteryLifeDays.toFixed(1)} days)

**Sleep time per cycle:** ${sleepTime.toFixed(1)} seconds
**Current savings ratio:** ${activeCurrent > 0 ? (activeCurrent / (sleepCurrent > 0 ? sleepCurrent : 0.001)).toFixed(0) : 0}× (deep sleep vs. active)`

Putting Numbers to It

Calculate the minimum solar panel size for year-round operation in Seattle (worst-case winter: 1.5 hours effective sunlight). The ESP32 node draws average 1.34mA at 3.7V.

Daily energy requirement: \[E_{daily} = I_{avg} \times V \times 24h = 1.34mA \times 3.7V \times 24h = 0.119Wh\]

Solar generation needed (with 70% charging efficiency): \[E_{solar} = \frac{0.119Wh}{0.70} = 0.17Wh\]

Winter panel output: \(P_{panel} \times 1.5h = 0.17Wh\), so \(P_{panel} = 0.113W\). With safety margin (2×): 0.23W minimum.

However, 3-day battery autonomy for cloudy periods requires storing \(3 \times 0.119Wh = 0.357Wh = 96mAh\) at 3.7V. Standard 18650 Li-ion (2500mAh) provides 9.25Wh – enough for 78 days of autonomy. A 2W panel ensures reliable winter charging even with week-long cloudy periods.

viewof avgCurrentSolar = Inputs.range([0.1, 10], {
  value: 1.34,
  step: 0.1,
  label: "Average Current Draw (mA):"
})

viewof voltageSystem = Inputs.range([3.0, 12.0], {
  value: 3.7,
  step: 0.1,
  label: "System Voltage (V):"
})

viewof sunlightHours = Inputs.range([0.5, 8], {
  value: 1.5,
  step: 0.5,
  label: "Effective Sunlight Hours (winter):"
})

viewof chargingEfficiency = Inputs.range([0.5, 0.9], {
  value: 0.70,
  step: 0.05,
  label: "Charging Efficiency:"
})

viewof autonomyDays = Inputs.range([1, 7], {
  value: 3,
  step: 1,
  label: "Battery Autonomy (cloudy days):"
})

dailyEnergyWh = (avgCurrentSolar / 1000) * voltageSystem * 24

energyRequiredWh = dailyEnergyWh / (chargingEfficiency > 0 ? chargingEfficiency : 0.01)

panelWatts = energyRequiredWh / (sunlightHours > 0 ? sunlightHours : 0.01)

panelWattsSafetyMargin = panelWatts * 2

batteryCapacityRequired = (dailyEnergyWh * autonomyDays) / (voltageSystem > 0 ? voltageSystem : 1)

md`### Solar Panel Sizing Calculator

**Daily Energy Consumption:** ${dailyEnergyWh.toFixed(3)} Wh
**Solar Energy Required (with ${(chargingEfficiency * 100).toFixed(0)}% efficiency):** ${energyRequiredWh.toFixed(3)} Wh
**Minimum Panel Size:** ${panelWatts.toFixed(2)} W
**Recommended (2× safety margin):** ${panelWattsSafetyMargin.toFixed(1)} W

**Battery Capacity for ${autonomyDays}-day autonomy:** ${(batteryCapacityRequired * 1000).toFixed(0)} mAh at ${voltageSystem.toFixed(1)}V
**Battery Energy Storage:** ${(batteryCapacityRequired * voltageSystem).toFixed(2)} Wh`

---

6.2.10 Cost Optimization Strategies

Budget-Conscious Choices:

• Prototyping: Buy from Adafruit/SparkFun for reliability and documentation ($30-50 kit)
• Small Production (10-50): Amazon for balance of cost/speed/support ($15-25 kit)
• Mass Production (100+): AliExpress bulk orders for 50-70% cost reduction ($8-12 kit)

Cost Comparison Example (10-node temperature monitoring) | Source | Per-Node Cost | Total Cost | Shipping | Lead Time | Quality | |——–|————–|———–|———-|———–|———| | Adafruit | $35 | $350 | $15 | 3-5 days | Excellent | | Amazon | $22 | $220 | Free | 2-3 days | Good | | AliExpress | $12 | $120 | $20 | 3-4 weeks | Variable |

Hidden Costs to Budget:

• Enclosures: $5-30 per node depending on IP rating
• Installation Labor: $20-100 per node for mounting, wiring
• Gateway/Hub: $50-200 (1 per site for LoRaWAN)
• Annual Maintenance: 10-15% of hardware cost
• Cloud/Data: $5-50/month depending on data volume

Quick Application-to-Hardware Guide

“I need to monitor…”

• Parking spaces, city-wide: ESP32 + PNI RM3100 magnetic sensor + LoRaWAN RFM95 + 5-year battery. Typical cost: about $90 per space. Range: 10 km. Life: 5-10 years.
• Soil moisture, farm with 50 locations: ESP32 + four capacitive soil sensors + LoRaWAN + solar. Typical cost: about $60 per node. Range: 5 km. Life: 5+ years with solar.
• Home energy usage: ESP32 + three SCT-013 current transformer clamps + Wi-Fi. Typical cost: about $45 total. Range: 50 m on Wi-Fi. Life: mains powered.
• Heart rate wearable: nRF52840 + MAX30102 + 250 mAh battery + BLE. Typical cost: about $35. Range: 10 m on BLE. Life: 7-14 days per charge.
• Machine vibration, 20 factory machines: STM32 + ADXL345 + Modbus RS485 + 24 V DC power. Typical cost: about $55 per machine. Range: 1 km wired. Life: industrial-grade.
• Neighbourhood air quality, 10 locations: ESP32 + PMS5003 + MH-Z19 + BME680 + solar. Typical cost: about $150 per node. Range: Wi-Fi or LoRaWAN. Life: 5+ years.

Not sure? Use the Sensor Selection Wizard below to get personalized recommendations!

Worked Example: Smart Parking Sensor Cost Analysis — Cloud vs. Embedded

Scenario: A city wants to deploy smart parking sensors across 500 parking spaces. Compare total cost of ownership (TCO) over 5 years for two architectures: camera-based with cloud AI vs. embedded magnetic sensor.

6.2.11 Option A: Camera + Cloud AI

Hardware per Space:

• IP camera (1080p): $80
• Edge gateway (1 per 10 spaces): $200 / 10 = $20
• Hardware Total: $100 per space

Connectivity & Cloud:

• 4G/LTE cellular data plan: $10/month per camera
• Cloud storage (video): $5/month per camera
• AI processing (object detection): $8/month per camera
• Monthly Cloud Cost: $23 per space

5-Year TCO per Space:

Hardware: $100
Installation: $50
Cloud (60 months): $23 × 60 = $1,380
Total: $1,530 per space

City-wide (500 spaces): 500 × $1,530 = $765,000

6.2.12 Option B: Embedded Magnetic Sensor

Hardware per Space:

• Magnetic sensor (PNI RM3100): $25
• ESP32 + LoRaWAN module: $13
• Battery (10-year CR123): $5
• Enclosure (IP67): $15
• Hardware Total: $58 per space

Infrastructure:

• LoRaWAN gateway (1 per 5 km²): $200
• Gateway internet (shared): $30/month for entire city
• Cloud storage (minimal JSON data): $0.50/month for entire city
• Monthly Cloud Cost: $30.50 total (≈$0.06 per space)

5-Year TCO per Space:

Hardware: $58
Installation: $25
Gateway (amortized): $200 / 500 = $0.40
Cloud (60 months): $0.06 × 60 = $3.60
Total: $87 per space

City-wide (500 spaces): 500 × $87 = $43,500

6.2.13 Cost Comparison:

Category	Camera + Cloud AI	Magnetic + LoRaWAN	Savings
Hardware	$50,000	$29,000	$21,000
Installation	$25,000	$12,500	$12,500
Connectivity (5 years)	$690,000	$1,830	$688,170
Total (5 years)	$765,000	$43,500	$721,500 (94%)
Annual Operating Cost	$138,000/year	$366/year	$137,634/year

6.2.14 Trade-Off Analysis:

Aspect	Camera + Cloud	Magnetic Sensor	Winner
Initial Cost	$150/space	$83/space	Magnetic (45% cheaper)
5-Year TCO	$1,530/space	$87/space	Magnetic (94% cheaper)
Power	Mains required	Battery (10 years)	Magnetic (no infrastructure)
Data Privacy	Video raises concerns	No PII	Magnetic (privacy-friendly)
Accuracy	95% (lighting dependent)	99% (detects metal)	Magnetic (more reliable)
Installation	$50 (wiring, mains)	$25 (surface mount)	Magnetic (simpler)
Maintenance	Camera cleaning, repairs	Battery swap (year 10)	Magnetic (minimal)
Latency	2-5 sec (cloud roundtrip)	<1 sec (local detection)	Magnetic (real-time)
Additional Features	Security surveillance	None	Camera (dual-use)

6.2.15 When to Choose Camera + Cloud:

• Parking lot also needs security surveillance (dual-use justifies cost)
• High-value locations where $1,530 per space is acceptable
• Existing 4G connectivity and mains power available
• City wants video analytics beyond parking (traffic flow, incident detection)

6.2.16 When to Choose Magnetic Sensor:

• Pure parking detection with no security requirements
• No mains power at parking spaces (battery-powered)
• Large deployments (>100 spaces) where TCO dominates decision
• Privacy-sensitive locations (residential streets, hospital visitor parking)
• Tight budget with 5-year planning horizon

6.2.17 Key Insights:

1. Connectivity dominates TCO for cloud-based solutions: $23/month × 60 months = $1,380 vs. $100 hardware cost
2. Battery-powered edge processing eliminates ongoing costs: Pay upfront, minimal operating expenses
3. LoRaWAN infrastructure is highly cost-effective: 1 gateway serves 500+ sensors over 5 km²
4. Embedded detection costs $0.06/month vs. cloud AI $8/month: 133× cheaper

Design Principle: For high-volume deployments with simple binary detection (occupied/empty), embedded sensors with low-power wireless crush cloud-based solutions on TCO. Reserve cameras + cloud AI for applications requiring complex visual analysis or dual-use cases.

viewof numSpaces = Inputs.range([50, 2000], {
  value: 500,
  step: 50,
  label: "Number of Parking Spaces:"
})

viewof yearsOperation = Inputs.range([1, 10], {
  value: 5,
  step: 1,
  label: "Years of Operation:"
})

// Camera + Cloud costs
cameraHardware = 80
cameraGateway = 200 / 10
cameraInstall = 50
cameraTotalHardware = cameraHardware + cameraGateway  // $100

cameraMonthlyData = 10
cameraMonthlyStorage = 5
cameraMonthlyAI = 8
cameraMonthlyCost = cameraMonthlyData + cameraMonthlyStorage + cameraMonthlyAI

cameraTCOPerSpace = cameraTotalHardware + cameraInstall + (cameraMonthlyCost * yearsOperation * 12)
cameraTCOTotal = numSpaces * cameraTCOPerSpace

// Magnetic sensor costs
magneticSensor = 25
magneticESP32 = 13
magneticBattery = 5
magneticEnclosure = 15
magneticInstall = 25
magneticTotalHardware = magneticSensor + magneticESP32 + magneticBattery + magneticEnclosure  // $58

magneticGateway = 200 / (numSpaces > 0 ? numSpaces : 1)
magneticMonthlyCloud = 30.50 / (numSpaces > 0 ? numSpaces : 1)

magneticTCOPerSpace = magneticTotalHardware + magneticInstall + magneticGateway + (magneticMonthlyCloud * yearsOperation * 12)
magneticTCOTotal = numSpaces * magneticTCOPerSpace

savingsTotal = cameraTCOTotal - magneticTCOTotal
savingsPercent = cameraTCOTotal > 0 ? ((savingsTotal / cameraTCOTotal) * 100) : 0

md`### TCO Comparison Calculator: Camera vs. Magnetic Sensor

#### Camera + Cloud AI
- **Hardware per space:** $${cameraTotalHardware.toFixed(2)}
- **Monthly cloud cost:** $${cameraMonthlyCost.toFixed(2)}/space
- **${yearsOperation}-year TCO per space:** $${cameraTCOPerSpace.toFixed(2)}
- **Total for ${numSpaces} spaces:** $${cameraTCOTotal.toLocaleString()}

#### Magnetic Sensor + LoRaWAN
- **Hardware per space:** $${magneticTotalHardware.toFixed(2)}
- **Gateway (amortized):** $${magneticGateway.toFixed(2)}/space
- **Monthly cloud cost:** $${magneticMonthlyCloud.toFixed(2)}/space
- **${yearsOperation}-year TCO per space:** $${magneticTCOPerSpace.toFixed(2)}
- **Total for ${numSpaces} spaces:** $${magneticTCOTotal.toLocaleString()}

#### Savings
**Total Savings:** $${savingsTotal.toLocaleString()} (${savingsPercent.toFixed(1)}%)
**Annual Savings:** $${(savingsTotal / yearsOperation).toLocaleString()}/year`

6.3 Diagrams

6.3.1 Sensor Application Domain Hierarchy

Figure 6.1: IoT Application Domains: Healthcare, Agriculture, Industrial, and Smart Home

Alternative View: Sensor Requirements by Application Criticality

This layered view categorizes sensors by criticality level - from safety-critical to convenience - helping you understand different reliability and accuracy requirements.

Figure 6.2: Criticality-based sensor selection: Safety-critical applications (medical, gas detection) demand certified sensors with redundancy. Operational applications (industrial monitoring) need high reliability. Informational applications (smart home) can use lower-cost consumer sensors. Match sensor grade to application criticality.

Figure 6.3: Sensor-to-domain mapping: Each IoT domain has characteristic sensors. Smart cities use ultrasonic (parking), cameras (traffic), and PM2.5 (air quality). Healthcare relies on PPG (heart rate), accelerometer (falls), and SpO2 (oxygen). Agriculture needs soil moisture, climate sensors, and GPS. Industrial IoT uses vibration, current, and flow sensors. Know your domain to select appropriate sensors.

6.3.2 Sensor Selection Decision Tree

Figure 6.4: Sensor Selection Decision Tree: Indoor vs Outdoor Deployment Paths

6.3.3 Data Flow Architecture

Figure 6.5: End-to-End IoT Data Flow: Sensors to Cloud to Application

6.3.4 Sensor Usage Statistics

Figure 6.6: IoT Sensor Market Distribution: Temperature Leads at 28%

6.3.5 Sensor Type Comparison by Application

Figure 6.7: Sensor Types by Application Domain: Smart Cities, Agriculture, Healthcare, Industrial

6.3.6 Sensor Deployment Lifecycle

Figure 6.8: Sensor Deployment Lifecycle: From Requirements to Scale and Optimization

6.4 Hands-On Labs

For Kids: Meet the Sensor Squad!

Choosing the right sensor is like choosing the right tool for a job!

6.4.1 The Sensor Squad Adventure: The Shopping Trip

The Sensor Squad had a mission: build a weather station for the school garden! But first, they needed to go shopping for the right parts.

Sammy the Sensor looked at the list. “We need a temperature sensor, a humidity sensor, and a rain gauge. But there are SO many choices!”

Max the Microcontroller said, “Let me help! First question: will this be INSIDE or OUTSIDE?”

“Outside!” everyone said. Max nodded. “Then we need waterproof sensors – ones rated IP65 or higher. That means they can handle rain and dust.”

Bella the Battery asked the important question: “How will we power it? There are no electrical outlets in the garden!”

Lila the LED had an idea. “Solar power! A small solar panel can charge a battery during the day, and we use just a tiny bit of power to take measurements.”

Max made the final decision. “We will use an ESP32 brain – it has Wi-Fi to send data to the classroom computer, and it is only $5! With a $12 solar panel, $5 temperature sensor, and $8 waterproof box, our total is about $30!”

Sammy was amazed. “That is less than a video game! And we get to watch the weather every day!”

6.4.2 Key Words for Kids

Word	What It Means
BOM	Bill of Materials – a shopping list for building something
IP Rating	How waterproof and dustproof something is (IP65 = very protected)
Solar Power	Using sunlight to make electricity
Budget	How much money you have to spend
Microcontroller	The tiny brain that reads sensors and sends data

Knowledge Check: Hardware Selection

Knowledge Check: Power Budget

Matching Quiz: Hardware Components and Their Roles

Ordering Quiz: IoT Hardware Design Process

Key Takeaway

Hardware selection for IoT deployments requires balancing multiple constraints: sensor accuracy vs. cost, power consumption vs. battery life, wireless range vs. data rate, and environmental protection vs. price. Always calculate total cost of ownership (hardware + installation + maintenance + connectivity + data storage) rather than just component cost, and pilot with a small deployment before scaling to avoid expensive mistakes.

Common Pitfalls

1. Using Breadboard Modules in Production Deployments

Prototype breakout boards with loose header pins are not suitable for deployed IoT nodes — vibration loosens connections, moisture causes corrosion, and pin headers are not rated for the connector cycles expected in maintained hardware. Design custom PCBs or use ruggedized modules with locking connectors for any deployment that needs to last more than a few weeks.

2. Ignoring Antenna Placement in Custom PCB Designs

Wireless connectivity performance degrades dramatically when the antenna trace is routed near ground planes, metal enclosures, or battery packs. On custom PCBs, follow the module manufacturer’s keep-out guidelines for antenna areas. In enclosures, position the module with the antenna facing away from metal surfaces.

3. Not Accounting for Sensor Module Current During Battery Sizing

Sensor module datasheets often specify operating current for the sensor IC alone, not for the complete module including its voltage regulator, onboard LED, and supporting circuits. Measure the actual current of the complete module assembly in-circuit at the expected supply voltage before calculating battery life.

4. Assuming USB Power Is Stable During IoT Development

USB host port current varies between 500 mA and 2.1 A depending on the port and negotiation. During Wi-Fi transmission spikes, ESP32 modules draw 500+ mA, which can exceed some USB port limits and cause brownout resets. Use a dedicated USB power adapter rated for at least 2 A during development, and design field hardware to use regulated DC supplies.

Label the Diagram

6.5 What’s Next

If you want to…	Read this
Learn which sensors and modules to use for specific application domains	Sensor Apps: Selection Wizard
Understand sensor application system architecture	Sensor Application Architecture
Practice building sensor applications in hands-on labs	Sensor Application Labs
Explore using mobile phones as a sensor hardware platform	Mobile Phone as a Sensor

Code Challenge

Order the Steps

Match the Concepts