3 Killer Thingino-Firmware Projects to Build in 2025
Ready to level up your embedded systems skills in 2025? Discover 3 killer firmware projects for the Thingino, from an Edge-AI camera to a Matter-native smart hub.
Alex Porter
An embedded systems engineer and tech writer passionate about making hardware accessible.
The new year is here, and for makers, that means a fresh slate of project ideas. If you’ve got a Thingino board burning a hole in your component drawer, you’re in the right place. The world of embedded systems is moving faster than ever, and 2025 is all about intelligence at the edge, seamless connectivity, and hyper-personal data.
Forget blinking an LED. We’re talking about projects that will seriously level up your firmware development skills and give you something genuinely impressive to show for it. These aren't just cool gadgets; they’re deep dives into C++, power management, and modern communication protocols.
Ready to fire up your IDE and soldering iron? Let’s dive into three killer Thingino-firmware projects to build in 2025.
Project 1: The Edge-AI Birdwatcher
Bird feeder cameras are a classic for a reason, but let’s bring this project into 2025. Instead of constantly streaming video to a cloud service for analysis, we’ll do the heavy lifting right on the device. This is the power of TinyML, or machine learning on microcontrollers.
Why It’s a Killer Project
This project teaches you the fundamentals of Edge AI. By running a lightweight machine learning model directly on a powerful board like the Thingino Pro, you create a device that’s faster, more power-efficient, and respects privacy. It only needs to connect to the internet to send you a notification (e.g., "A Northern Cardinal is at the feeder!") rather than streaming gigabytes of data. This is the future of smart, autonomous IoT devices.
Core Components
- Thingino Pro 2: You'll need the processing power and RAM for the ML model.
- Camera Module: An ArduCam or ESP32-CAM compatible module works great.
- PIR Sensor: To wake the device from deep sleep when there's motion.
- Solar Panel & LiPo Battery: For a truly wireless, set-and-forget installation.
- 3D Printed Enclosure: To keep everything weatherproof.
The Firmware Challenge
This is where your skills will shine. Your C++ code will be the brain of the entire operation. You'll need to:
- Implement Aggressive Power Management: The Thingino should be in deep sleep for 99% of the time, drawing microamps of current. The PIR sensor will trigger an interrupt to wake it up.
- Interface with the Camera: Write the driver code to capture a high-quality image upon wakeup.
- Integrate a TensorFlow Lite Micro Model: This is the core of the project. You'll take a pre-trained bird classification model (like one from MobileNetV2, optimized for microcontrollers) and write the firmware to load the image data into it, run the inference, and parse the results.
- Handle Connectivity: Based on the inference result (e.g., if a bird is identified with >90% confidence), your firmware will wake the Wi-Fi module, connect to your network, send a notification via MQTT or a webhook, and then immediately go back to sleep.
Project 2: The Matter-Native Smart Energy & Air Quality Hub
The smart home is finally getting standardized with Matter. Instead of building another proprietary Wi-Fi gadget, let's build a device that speaks the new universal language of IoT. This hub won’t just monitor; it will be a central node for understanding and improving your home environment.
Why It’s a Killer Project
Building a Matter device from scratch is the best way to understand the next generation of smart home technology. You’re not just learning a protocol; you’re learning how devices will securely and reliably interoperate for the next decade. This project combines practical home utility (saving energy, breathing cleaner air) with cutting-edge connectivity skills.
Core Components
- Thingino Thread Edition: A board with built-in 802.15.4 radio support for Thread.
- SCT-013 Current Sensor: A non-invasive clamp to measure whole-home energy usage.
- SCD4x CO2 Sensor: For accurate carbon dioxide monitoring.
- BME688 or SGP40 Sensor: To measure Volatile Organic Compounds (VOCs) and overall air quality.
- PM2.5 Sensor (e.g., PMS5003): To detect fine particulate matter.
The Firmware Challenge
This is a networking and sensor integration beast. The firmware is responsible for making everything work together seamlessly within the Matter ecosystem.
- Implement the Matter Device Data Model: You'll define your device in firmware as a composite of multiple Matter device types: a power meter, an air quality sensor, a CO2 sensor, etc. This involves setting up ZAP (ZCL Advanced Platform) templates and generating the right data structures.
- Manage the Thread Network Stack: Your firmware will handle joining the Thread mesh network, maintaining its connection, and routing messages. You'll work directly with the OpenThread stack.
- Poll Multiple I2C/UART Sensors: You'll write robust, non-blocking code to periodically read data from all your different sensors, handle potential errors, and store the values.
- Expose Attributes via Matter: The core task is to link your sensor readings to the Matter attributes. When a Matter controller (like your phone or smart speaker) asks for the CO2 level, your firmware needs to provide the latest value from the SCD4x sensor in the correct format.
Project 3: The Personal Enviro-Monitor Wearable
We carry devices that track our steps and heart rate, but what about the environment immediately around us? This project is a small, wearable device that gives you a real-time score of your personal environmental exposure, from UV rays to air pollution.
Why It’s a Killer Project
This project is a masterclass in power optimization and sensor fusion. Making a battery-powered device that lasts for days while constantly sensing and communicating is a huge challenge. You'll also learn how to take raw data from multiple sources and combine it into a single, meaningful metric—a skill that’s invaluable in any data-driven field.
Core Components
- Thingino Nano BLE: A tiny, low-power board with Bluetooth Low Energy.
- VEML6075 UV Sensor: To measure UVA/UVB index.
- MICS-4514 or similar Gas Sensor: To detect pollutants like NO2 and CO.
- Small E-Ink Display (1-inch): For ultra-low-power status updates.
- Tiny LiPo Battery & Charger: The smaller, the better.
- Compact 3D Printed Case: To make it a true wearable.
The Firmware Challenge
Efficiency is the name of the game here. Every line of code impacts battery life.
- Ultra-Low-Power Architecture: Your main loop should be empty. Everything must be interrupt-driven or handled by low-power timers. You'll spend most of your time managing sleep modes and deciding which peripherals can be powered down.
- Bluetooth Low Energy (BLE) Services: You’ll design a custom BLE service in your firmware with characteristics for the UV index, pollution level, and the final “Enviro-Score.” Your phone will connect to this service to read the data.
- Sensor Fusion Algorithm: This is the creative part. How do you combine a UV index of 8, an NO2 reading of 50 ppb, and a temperature of 30°C into one score? You'll write the logic in C++ to weigh these factors and produce a simple 1-100 score.
- E-Ink Display Driver: E-Ink displays are fantastic for low power but have slow refresh rates and require specific update sequences. Your firmware will need to handle this, only updating the display when a value changes significantly to save power.
Get Building!
These projects are more than just weekend builds; they are deep dives into the skills that define modern firmware engineering. Whether you choose to tackle Edge AI, next-gen connectivity, or ultra-low-power design, you'll come out the other side with a much stronger understanding of what it takes to build professional-grade embedded devices.
So, which one are you going to build first? Let us know in the comments below!