Project Aura: From Idea to Reality

February 6, 2025

Welcome to the first official project log for Project Aura! I'm excited to finally share this journey with the Hackaday community.

How It All Started

This project didn't begin with a business plan or a grand vision. It started with a simple observation: we obsess over tracking our steps, calories, and sleep — but we completely ignore the air we breathe 20,000 times a day.

I noticed my productivity dropping in the afternoons. Headaches. Brain fog. I blamed stress and poor sleep. Then I bought a cheap CO₂ monitor and discovered my home office was hitting 1800+ ppm regularly — well into the "cognitive impairment" zone according to research.

That was my wake-up call.

Why Build Instead of Buy?

I looked at commercial air quality monitors. They all had the same problems:

• Closed-source firmware (no way to verify accuracy)
• Cloud-dependent (what happens when the company shuts down?)
• Limited sensor selection (usually just PM2.5 and CO₂)
• Expensive ($200-500 for decent units)
• Poor Home Assistant integration (if any)

As a maker, I thought: "I can build something better. And I can make it open source so others can too."

Design Philosophy

From day one, I had three non-negotiable principles:

1. No soldering required — This needed to be accessible to anyone, not just electronics experts
2. Industrial-grade sensors — No cheap DHT11/MQ-135 clones. Real Sensirion sensors or nothing.
3. Professional UX — No "Arduino project" aesthetic. It needed to look and feel like a commercial product.

The Hardware Journey

I went through three major iterations:

Version 1: ESP32 DevKit + 2.8" resistive touchscreen + breadboard mess. It worked, but looked terrible. Cable management was a nightmare.

Version 2: Custom PCB + 3.5" capacitive display. Better, but still felt like a prototype. Manufacturing costs would've been too high for a DIY project.

Version 3 (Current): Waveshare ESP32-S3-Touch-LCD-4.3 + Grove/STEMMA QT ecosystem. This was the breakthrough. The integrated display eliminated cable spaghetti, and standardized connectors made assembly trivial.

The Sensirion SEN66 Decision

This was the most expensive component choice (around $60), and I debated it for weeks.

Cheaper alternatives like SGP30 + SCD30 + PMS5003 would've saved $30-40. But they have problems:

• Separate sensors = more cables, more space, more complexity
• No VOC/NOx indices (just raw resistance values that drift over time)
• Optical PM sensors contaminate quickly without proper airflow design

The SEN66 uses Sensirion's Sheath Flow technology — it actively pulls air through a protected channel, keeping the optics clean for years. The VOC and NOx algorithms are calibrated against human perception studies.

In the end, reliability won. This device needs to work accurately for years, not months.

The UI Challenge

Getting LVGL running smoothly on ESP32 was harder than I expected. The 800×480 display needs a ton of RAM for the framebuffer. I had to:

• Use PSRAM aggressively
• Optimize image assets (convert to indexed color)
• Implement smart buffering to avoid tearing
• Pack the Theme Studio web interface into PROGMEM

But the result? smooth gradients, and zero lag when changing themes. Worth every hour of optimization.

Home Assistant Integration: The Killer Feature

I knew from the start this needed seamless Home Assistant integration. Manual YAML configuration is a dealbreaker for most users.

MQTT Auto-Discovery was the answer. The device publishes its entire configuration on startup, and Home Assistant automatically creates 20+ entities. No configuration files. No manual setup. Just works.

I spent weeks perfecting the entity metadata (device classes, units, icons) so everything looks native in HA dashboards.

Going Open Source

I wrestled with this decision. Should I keep the firmware closed and try to monetize it?

Ultimately, transparency won. Air quality affects health. People deserve to know exactly how their data is being collected and processed. Open source builds trust.

The crowdfunding model on MakerWorld lets me monetize the design work (enclosure files + assembly guide) while keeping the firmware free and inspectable.

Current Status & What's Next

The prototype is fully functional and has been running 24/7 in my home office for the past three weeks. Rock solid.

Completed:

• ✅ Hardware design finalized
• ✅ Firmware feature-complete
• ✅ MQTT/Home Assistant integration tested
• ✅ Multi-language UI implemented
• ✅ Theme Studio web interface working
• ✅ Assembly guide in progress (80% done)

Coming Soon:

• 📝 Final assembly PDF with high-res photos
• 📦 Alternative enclosure designs (wall-mount variant, compact version)
• 🔧 Web-based firmware updater (OTA without PlatformIO)
• 📊 Optional cloud dashboard for multi-device tracking

Community Feedback Welcome!

This project is now live on MakerWorld and has already received incredible support from the community. Over 1,000 backers in just a few days!

I'm actively gathering feedback on:

• Which sensors are most important to you?
• Would you prefer a smaller enclosure (no formaldehyde sensor)?
• Are there other air quality metrics you'd like to see?
• What Home Assistant automations would you build with this data?

Feel free to drop questions or suggestions in the comments. This is a community-driven project, and your input directly shapes future development.

Resources

• GitHub: github.com/21cncstudio/project_aura
• MakerWorld Campaign: https://makerworld.com/en/crowdfunding/159-project-aura-make-the-invisible-visible
• Live Demo Video:

---

Stay tuned!

— Volodymyr Papush