12 Common Hardware Platforms

12.1 Learning Objectives

By the end of this chapter, you will be able to:

Compare Arduino, ESP32, Raspberry Pi, STM32, and Particle platforms across power, cost, connectivity, and ecosystem dimensions
Justify platform selection for specific IoT project requirements using quantitative criteria such as battery life and total cost of ownership
Operate interactive simulators to prototype and validate firmware on different platforms before purchasing hardware
Assess the strengths, limitations, and production scalability of each platform family

Key Concepts

Prototyping Platform: Standardised development board ecosystem (Arduino, Raspberry Pi, ESP32) with consistent pinout, libraries, and community support.
Arduino: Beginner-friendly MCU ecosystem with standardised shields and a large library ecosystem, ideal for rapid iteration.
ESP32: Dual-core MCU with Wi-Fi and Bluetooth, the dominant prototyping platform for connected IoT devices at $5 per module.
Raspberry Pi: Linux-based single-board computer used for IoT gateways, edge AI, and applications requiring full OS support.
MicroPython: Python 3 subset running directly on microcontrollers, enabling rapid scripting without C/C++ compilation.
PlatformIO: Cross-platform embedded development environment supporting 900+ boards with dependency management and CI integration.
Community Ecosystem: Libraries, tutorials, and forum support that dramatically accelerates development on popular prototyping platforms.

For Beginners: Common Hardware Platforms

Prototyping is building rough, working versions of your IoT device to test ideas quickly and cheaply. Think of it like building a model airplane before constructing the real thing – a prototype reveals problems when they are still easy and inexpensive to fix. Modern prototyping tools make it possible to go from idea to working device in days rather than months.

Sensor Squad: Choosing Your Platform

“Picking the right hardware platform is like choosing the right vehicle for a journey!” said Max the Microcontroller. “An Arduino Uno is like a bicycle – simple, reliable, and great for short trips. An ESP32 is like a car – it has Wi-Fi and Bluetooth built in, so it can go places. A Raspberry Pi is like a truck – powerful enough to carry heavy loads like machine learning and video processing.”

Sammy the Sensor was overwhelmed. “How do I choose?” Max broke it down. “Ask yourself three questions. Do you need wireless connectivity? Go with ESP32. Do you need to run Linux or process images? Get a Raspberry Pi. Is your project simple with just a few sensors? An Arduino is perfect and costs less than five dollars.”

Lila the LED shared a tip. “For IoT prototyping, the ESP32 is the most popular choice because it includes Wi-Fi and Bluetooth at a low price. You can even simulate it online with Wokwi before buying real hardware!” Bella the Battery added her perspective. “From a power standpoint, Arduino and ESP32 can run on batteries for months. The Raspberry Pi needs a constant power supply – it draws way too much current for battery operation. Always consider power requirements when choosing your platform!”

Interactive: Hardware Selection Optimizer

12.2 Arduino Family

12.2.1 Arduino Uno

MCU: ATmega328P (8-bit AVR)
Speed: 16 MHz
Memory: 32 KB Flash, 2 KB RAM
I/O: 14 digital, 6 analog
Best For: Learning, simple projects, hobbyist prototypes

12.2.2 Arduino Mega

MCU: ATmega2560
Memory: 256 KB Flash, 8 KB RAM
I/O: 54 digital, 16 analog
Best For: Projects requiring many I/O pins

12.2.3 Arduino Nano/Mini

Compact form factors
Same MCU as Uno
Best For: Space-constrained projects

12.2.4 Arduino Due

MCU: ARM Cortex-M3 (32-bit)
Speed: 84 MHz
Memory: 512 KB Flash, 96 KB RAM
Best For: Computationally intensive applications

12.2.5 Arduino Strengths and Limitations

Strengths:

Beginner-friendly IDE and ecosystem
Massive community and library support
Standardized form factor (shields)
Abundant tutorials and examples

Limitations:

Limited processing power
No built-in Wi-Fi/Bluetooth (except Wi-Fi-enabled variants)
5V logic (compatibility issues with 3.3V devices)

12.2.6 Knowledge Check

12.3 ESP32/ESP8266 Family

12.3.1 ESP32

MCU: Dual-core Xtensa LX6, 240 MHz
Memory: 520 KB RAM, 4 MB Flash (typical)
Connectivity: Wi-Fi 802.11 b/g/n, Bluetooth Classic + BLE
Peripherals: Rich (ADC, DAC, touch sensors, SPI, I2C, UART)
Power Modes: Deep sleep < 10 uA

12.3.2 ESP8266

MCU: Single-core Xtensa L106, 80/160 MHz
Connectivity: Wi-Fi 802.11 b/g/n
Lower cost and simpler than ESP32

12.3.3 Strengths and Limitations

Strengths:

Built-in Wi-Fi/Bluetooth
Low cost ($2-10)
Low power with deep sleep
Arduino IDE compatible
Active community

Limitations:

Fewer I/O pins than Arduino Mega
3.3V logic only
Wi-Fi stack consumes resources

Best For:

Wi-Fi-connected sensors
Home automation
IoT gateways
Battery-powered Wi-Fi devices

12.3.4 Knowledge Check

12.4 Interactive Simulator: Arduino Blink

Arduino Uno - Blink (The “Hello World” of Hardware)

What This Simulates: Classic Arduino Uno LED blink program - the first project for every maker

Programming Fundamentals:

digitalWrite() function
delay() timing
setup() and loop()
Pin modes

Hardware Fundamentals:

GPIO pins (digital output)
LED forward voltage
Current limiting resistor
Active HIGH/LOW

How to Use:

Click Start Simulation
Watch built-in LED on pin 13 blink every second
Modify delay() values to change blink rate
Try different pins (2-13) with external LED

12.5 Raspberry Pi Family

12.5.1 Raspberry Pi 4 Model B

Processor: Quad-core ARM Cortex-A72, 1.5 GHz
RAM: 1/2/4/8 GB options
Connectivity: Gigabit Ethernet, Wi-Fi, Bluetooth
Interfaces: USB 3.0, HDMI, GPIO header
OS: Raspberry Pi OS (Linux), Ubuntu, Windows IoT

12.5.2 Raspberry Pi Zero/Zero W

Compact, low-cost variant
Single-core ARM11, 1 GHz
512 MB RAM
Zero W adds Wi-Fi/Bluetooth

12.5.3 Raspberry Pi Pico

MCU-class board (not Linux-capable)
RP2040 dual-core Cortex-M0+, 133 MHz
264 KB RAM
Best For: MCU projects needing custom silicon

12.5.4 Strengths and Limitations

Strengths:

Full Linux environment
Rich software ecosystem
Powerful processing for edge computing
Extensive peripherals
Large community

Limitations:

Higher power consumption (2-7W)
Not real-time (Linux has scheduling latency)
Larger form factor
Requires SD card for boot

Best For:

Edge gateways
Video processing
Machine learning inference
Complex IoT applications

12.5.5 Knowledge Check

12.6 Other Platforms

12.6.1 BeagleBone Family

BeagleBone Black:

Processor: ARM Cortex-A8, 1 GHz
RAM: 512 MB DDR3
Storage: 4 GB eMMC onboard
PRUs: Programmable Real-time Units for deterministic I/O
OS: Debian Linux

In 60 Seconds

This chapter covers common hardware platforms, explaining the core concepts, practical design decisions, and common pitfalls that IoT practitioners need to build effective, reliable connected systems.

Strengths:

More I/O than Raspberry Pi (65 GPIO)
Real-time capable with PRUs
Industrial temperature variants
Open-source hardware

Best For:

Industrial applications
Real-time I/O with Linux
Robotics and automation

12.6.2 STM32 Nucleo/Discovery Boards

Nucleo Boards:

Various STM32 MCU options (Cortex-M0/M3/M4/M7)
Arduino-compatible headers
ST-Link debugger onboard
Professional development environment (STM32CubeIDE)

Strengths:

Industrial-grade MCUs
Low power optimization
Professional toolchain
HAL libraries

Best For:

Production-intent prototypes
Low-power battery applications
Professional embedded development

12.6.3 Particle Platform

Particle Photon/Argon/Boron:

MCU with integrated Wi-Fi or cellular connectivity
Cloud platform included
Over-the-air (OTA) updates
Cellular data plans for Boron

Strengths:

End-to-end IoT platform
Easy cloud connectivity
Built-in device management
Rapid prototyping for connected devices

Best For:

Connected products requiring cloud integration
Fleet management applications
Rapid IoT prototyping

12.7 Worked Example: Choosing a Platform for a Smart Beehive Monitor

A beekeeper cooperative wants to monitor 200 beehives across 15 apiaries. Each hive needs temperature, humidity, and weight sensors with hourly readings, transmitted to a cloud dashboard. Budget: $50 per hive, battery life: 1 year minimum, range to nearest gateway: up to 500 meters.

Step 1: Eliminate non-contenders

Requirement	Arduino Uno	ESP32	Raspberry Pi 4	STM32L4 Nucleo	Particle Boron
Built-in wireless	No	Wi-Fi+BLE	Wi-Fi+ETH	No	Cellular+BLE
Deep sleep <50 uA	N/A	10 uA	N/A (600 mA idle)	1.1 uA	3.5 uA
Unit cost	$25	$5-10	$55	$15	$40 + data plan
Fits $50 budget (with sensors)	Tight	Yes	No	Yes	No (recurring cost)

Raspberry Pi is eliminated immediately – 600 mA idle draws would drain any reasonable battery in hours. Particle Boron’s $3/month cellular plan adds $36/year recurring cost per hive. Arduino Uno requires a separate wireless module ($10-15), pushing cost to $40-50 without sensors.

Step 2: Deep comparison of finalists

Factor	ESP32	STM32L4 + LoRa module
Unit cost (board only)	$5	$15 + $12 (RFM95W) = $27
Wireless range	50-100 m (Wi-Fi)	2-5 km (LoRa)
Gateway requirement	Wi-Fi AP at each apiary ($30 x 15 = $450)	3 LoRa gateways ($300 x 3 = $900)
Power per transmission	200 mA x 2s = 0.11 mAh	120 mA x 0.1s = 0.003 mAh
Sleep current	10 uA	1.1 uA
Average current (hourly TX)	121 uA	4.4 uA
Battery life (3000 mAh)	3.4 months	2.8 years
Development ecosystem	Arduino IDE, vast libraries	STM32CubeIDE, steeper learning curve
Community support	Massive (hobbyist + professional)	Professional-focused

Step 3: Total cost of ownership (200 hives, 3 years)

Cost Item	ESP32 + Wi-Fi APs	STM32L4 + LoRa
Sensor nodes (200 x board + sensors)	200 x $35 = $7,000	200 x $47 = $9,400
Gateways/infrastructure	15 x $30 = $450	3 x $300 = $900
Battery replacements (3 years)	2,143 x $2 = $4,286	200 x $2 = $400
Cellular/connectivity	15 x $30/mo x 36 = $16,200 (cellular backhaul)	3 x $10/mo x 36 = $1,080 (TTN)
3-year total	$27,936	$11,780

Decision: STM32L4 + LoRa wins despite higher per-node hardware cost ($47 vs $35). The 27× lower average current (4.4 µA vs 121 µA) dramatically reduces battery replacements (200 vs 2,143 over 3 years), and LoRa’s 2-5 km range means 3 gateways replace 15 Wi-Fi access points. The $16,156 savings over 3 years covers the extra development time from STM32’s steeper learning curve.

Putting Numbers to It

The 27× average current difference between platforms determines battery lifespan. For hourly transmissions with 3000 mAh capacity:

ESP32: Average current for hourly transmissions:

\[ I_{\text{avg}} = \frac{(10\ \mu\text{A} \times 3598\text{ s}) + (200\text{ mA} \times 2\text{ s})}{3600\text{ s}} = \frac{35,980\ \mu\text{A·s} + 400,000\ \mu\text{A·s}}{3600\text{ s}} = 121\ \mu\text{A} \]

Battery life:

\[ t_{\text{life}} = \frac{3000\text{ mAh}}{0.121\text{ mA}} = 24,793\text{ hours} = 3.4\text{ months} \]

STM32L4: Average current for hourly transmissions:

\[ I_{\text{avg}} = \frac{(1.1\ \mu\text{A} \times 3599.9\text{ s}) + (120\text{ mA} \times 0.1\text{ s})}{3600\text{ s}} = \frac{3,960\ \mu\text{A·s} + 12,000\ \mu\text{A·s}}{3600\text{ s}} = 4.4\ \mu\text{A} \]

Battery life:

\[ t_{\text{life}} = \frac{3000\text{ mAh}}{0.0044\text{ mA}} = 681,818\text{ hours} = 2.8\text{ years} \]

At 200 hives, the ESP32 fleet requires $200 \times (3 \text{ years} / 0.28 \text{ years}) = 2,143$ battery replacements over 3 years at $2 per unit = $4,286 labor + parts. The STM32L4 fleet needs one replacement per hive ($200 \times 1 = 200$ batteries at $2 = $400), saving $3,886.

When would ESP32 win instead? If apiaries had existing Wi-Fi (e.g., near farmhouses), eliminating gateway costs. Or if readings were needed every minute rather than hourly – Wi-Fi’s higher bandwidth handles larger payloads more efficiently than LoRa’s 242-byte limit.

Interactive Calculator: Battery Life & TCO Comparison

Explore how transmission frequency and sleep current affect battery life and total cost of ownership across platforms.

Show code

viewof platform = Inputs.select(["ESP32", "STM32L4"], {label: "Platform", value: "ESP32"})
viewof sleep_current = Inputs.range([1, 200], {value: platform === "ESP32" ? 10 : 1.1, step: 0.1, label: "Sleep current (µA)"})
viewof tx_current = Inputs.range([50, 300], {value: platform === "ESP32" ? 200 : 120, step: 10, label: "TX current (mA)"})
viewof tx_duration = Inputs.range([0.05, 5], {value: platform === "ESP32" ? 2 : 0.1, step: 0.05, label: "TX duration (seconds)"})
viewof tx_interval = Inputs.range([1, 60], {value: 60, step: 1, label: "TX interval (minutes)"})
viewof battery_capacity = Inputs.range([1000, 5000], {value: 3000, step: 100, label: "Battery capacity (mAh)"})
viewof num_devices = Inputs.range([10, 500], {value: 200, step: 10, label: "Number of devices"})
viewof project_years = Inputs.range([1, 5], {value: 3, step: 1, label: "Project duration (years)"})

Show code

battery_calcs = {
  const interval_seconds = tx_interval * 60;
  const sleep_time = interval_seconds - tx_duration;
  const avg_current_ua = (sleep_current * sleep_time + tx_current * 1000 * tx_duration) / interval_seconds;
  const battery_life_hours = battery_capacity / (avg_current_ua / 1000);
  const battery_life_months = battery_life_hours / 730; // 730 hours per month average
  const battery_life_years = battery_life_months / 12;
  const replacements_per_device = Math.max(0, Math.floor(project_years / battery_life_years));
  const total_replacements = num_devices * replacements_per_device;
  const replacement_cost = total_replacements * 2; // $2 per battery

  return {
    avg_current_ua: avg_current_ua.toFixed(1),
    battery_life_hours: battery_life_hours.toFixed(0),
    battery_life_months: battery_life_months.toFixed(1),
    battery_life_years: battery_life_years.toFixed(2),
    replacements_per_device: replacements_per_device,
    total_replacements: total_replacements,
    replacement_cost: replacement_cost
  };
}

Show code

html`<div style="background: var(--bs-light, #f8f9fa); padding: 1.5rem; border-radius: 8px; border-left: 4px solid #3498DB; margin-top: 1rem;">
<table style="width:100%; border-collapse:collapse;">
<tr style="background: #2C3E50; color: white;">
  <th style="padding:8px; text-align:left;">Metric</th>
  <th style="padding:8px; text-align:right;">Value</th>
</tr>
<tr style="border-bottom: 1px solid #ddd;">
  <td style="padding:8px;"><strong>Average Current</strong></td>
  <td style="padding:8px; text-align:right;">${battery_calcs.avg_current_ua} µA</td>
</tr>
<tr style="border-bottom: 1px solid #ddd; background: #f8f9fa;">
  <td style="padding:8px;"><strong>Battery Life</strong></td>
  <td style="padding:8px; text-align:right;">${battery_calcs.battery_life_months} months (${battery_calcs.battery_life_years} years)</td>
</tr>
<tr style="border-bottom: 1px solid #ddd;">
  <td style="padding:8px;"><strong>Replacements per Device</strong></td>
  <td style="padding:8px; text-align:right;">${battery_calcs.replacements_per_device}</td>
</tr>
<tr style="border-bottom: 1px solid #ddd; background: #f8f9fa;">
  <td style="padding:8px;"><strong>Total Replacements (${num_devices} devices)</strong></td>
  <td style="padding:8px; text-align:right;">${battery_calcs.total_replacements.toLocaleString()}</td>
</tr>
<tr style="background: #16A085; color: white;">
  <td style="padding:8px;"><strong>Total Battery Cost (${project_years} years)</strong></td>
  <td style="padding:8px; text-align:right; font-size: 1.2em;"><strong>$${battery_calcs.replacement_cost.toLocaleString()}</strong></td>
</tr>
</table>
<p style="margin-top: 1rem; font-size: 0.9em; color: #7F8C8D;"><em>Note: Assumes $2 per battery replacement including labor. Does not include initial batteries.</em></p>
</div>`

Try This:

Switch between ESP32 and STM32L4 to see the default parameters
Change transmission interval from hourly (60 min) to every 10 minutes - watch battery life plummet
Compare a 200-device deployment over 3 years with different platforms
Explore the impact of improving sleep current from 10 µA to 5 µA on ESP32

12.8 Platform Comparison Summary

Platform	Processor	Connectivity	Power	Cost	Best For
Arduino Uno	8-bit AVR 16MHz	None	45mA	$25	Learning
Arduino Mega	8-bit AVR 16MHz	None	50mA	$45	Many I/O
ESP32	Dual-core 240MHz	Wi-Fi + BLE	10uA sleep	$5-10	IoT
ESP8266	Single-core 80MHz	Wi-Fi	10uA sleep	$3-5	Simple IoT
Raspberry Pi 4	Quad-core 1.5GHz	Wi-Fi + ETH	600mA+	$35-75	Edge compute
STM32 Nucleo	Cortex-M various	Add modules	1uA sleep	$10-25	Professional

Interactive Quiz: Match Platforms to Use Cases

🏷️ Label the Diagram

💻 Code Challenge

📝 Order the Steps