We’re excited to launch our next community-driven Project Challenge right here at Axelera AI: Smarter Spaces. Following the success of Pioneer 10, this new challenge invites you to think smaller, closer, and more personal as you bring edge AI into your immediate surroundings. AI should live in your space, not in the cloud!
As developers, makers, engineers, IoT enthusiasts and AI tinkerers, your mission is simple:
Embed edge intelligence into the places people live, work, visit and play, while keeping data local, private, and out of the cloud.
What is Smarter Spaces?
In this challenge, you’ll build an AI-powered system for your home, workspace, or building that can:
-
Sense and understand its environment via camera(s) or other sensors
-
React meaningfully (e.g. control devices, send alerts, trigger automations, detect objects or events)
-
Operate fully offline (even if internet connectivity is available)
-
Maintain privacy, reliability, and resilience
Some possible directions you could consider:
-
🛡️ Smart security or package detection (without sending video to the cloud)
-
🦺 Safety & compliance monitoring in workshop or shared spaces
-
💡 Context-aware lighting, HVAC, or appliance control
-
✋ Gesture- or occupancy-driven UI or interfaces
-
🏠 Home, workplace or life automation through edge AI
As always, you’re encouraged to get super creative, but with an eye on realism and edge feasibility. After all, if you’re selected, you’ll actually have to build your project and make it work!
What You’ll Receive (The Kit)
We want you to hit the ground running while still having freedom to customise your build. The free edge AI kit includes:
-
Axelera Metis M.2 (to accelerate vision AI)
-
Orange Pi 5 Plus host system (provided via our awesome partner in this project challenge, Orange Pi)
-
4× Cameras (RTSP/USB 1080p, ONVIF-compatible)
-
2 × Sonoff CAM-S2
-
2 × Sonoff CAM-PT2
-
-
HDMI display (7–10”) for local feedback or monitoring
-
Networking & accessories
-
WiFi board (for optional connectivity)
-
Mini-router
-
Mounts, cables, power supplies as needed
-
This mix is designed to let you iterate fast on vision + edge. You’re free to add your own sensors, modules, software, models or hardware as needed.
Even though your project will need to fully operate offline, you might notice that WiFi options are included. This is in case you may want to set up a local WiFi network to reduce reliance on cabling, or for simpler data exchange. So wireless data is totally fine (great, in fact!) but even that wireless network must be local.
So your project can be connected to the internet, but it also needs to be fully operational when it’s offline, too.
We’re proud to partner with Orange Pi on this, as their boards bring amazing capabilities for edge systems and form a perfect match with Metis and our software stack. If you’ve spent any time here on the Axelera AI community, you’ll already know that Orange Pi + Metis is a great combo.
What’s In It for You, And the Community
Smarter Spaces is more than an IoT or automation contest. It’s a chance for you to:
-
Build real, deployable AI systems and add them to your portfolio. You keep your IP.
-
Explore offline-first vision pipelines using Voyager SDK (that’s where the AI industry is moving, after all)
-
Share your learnings, code, and solutions with other makers
-
Capture attention in the edge AI, IoT, and maker worlds
-
Win prizes, recognition, and experience for your next projects
-
🥇First place: €1,000
-
🥈Runners up: 2 x €500
-
From the community side, we’ll be curating and showcasing standout projects, producing spotlight case studies, and amplifying your work through our channels and partner networks (including Orange Pi’s awesome and vast audience).
How It Works & Timeline
-
Pitch your idea
Describe your concept: what you’ll sense, how you’ll react, which hardware/software you’ll use, and why it matters.
-
Selection & distribution
We’ll choose a 10 proposals and send out kits to those selected.
-
Build & iterate
Develop your project over a period of X weeks, posting regular build updates, sharing challenges, and getting help and feedback.
-
Submit & demo
You’ll provide a demo video, project write-up, and code repo for your finished project, and your chance for the grand prizes.
-
Judging & awards
We’ll evaluate for originality, technical merit, completeness, presentation, and edge viability.
Submissions are now closed, but stay tuned to find out which projects have been selected soon!
