Mary Karapetyan@mekarape

LockLens is a pay-per-inference AI vision platform that enables any Linux-based robot with a serial motor controller to follow any target specified in plain English. No fiducial markers, no physical setup, no retraining required. The reference implementation runs on a custom hoverboard conversion with 8.5 inch hub motors, Flipsky VESC motor controllers connected via CAN bus, a 36V battery system with onboard charging, and a Raspberry Pi 4B running the full inference stack. A 5 inch touchscreen dashboard displays the live camera feed, bounding box overlay, latency, confidence, frame count, and cost in real time. The AI pipeline sends camera frames from the Pi to a Qwen2.5-VL-7B-Instruct model served by vLLM on an AMD MI300X GPU via AMD Developer Cloud. The model returns normalized bounding box coordinates which feed into a proportional controller driving the motors. Each inference call is metered via X402 micropayments, fractions of a cent per frame, no subscription, no charge when idle. Built solo for the AMD Developer Hackathon 2026, Track 3: Launch and Fund Your Startup.

Grab your ideas, spin up compute, and start building what’s next! Start with $100 in AMD Developer Cloud credits and begin building right away. Get access to hands-on tutorials, workshops, and curated learning resources to help you build faster and go further. To unlock credits and join the hackathon, sign up for the AMD AI Developer Program. 🤝 Go solo or team up — ship something real. 💰 $21,500+ total prize pool & 1 × AMD Radeon™ AI PRO R9700 GPU up for grabs. 📍 On-site Venue (May 9-10): MindsDB SF AI Collective 3154 17th St, San Francisco, California, USA 📍On-site participation is by invitation only. Travel and accommodation expenses will not be covered. 🧑‍💻 Developers, builders, and innovators - bring your boldest ideas and let’s turn them into real AI systems on AMD.