A robot that dreams itself better — a self-learning system for the Arduino Ventuno Q. Robots leave the factory and never learn again. REM PhysicalAI changes that. REM PhysicalAI is a self-improvement system that installs on a robot. Give a robot a way to act, a camera to see, and a computer to think — here, an Arduino Ventuno Q running ExecuTorch, 100% on-device — and it learns by itself: it commands an action, watches what happened, and relates the two. When nobody is using it, it dreams. In its dream it practices in simulation — standing, walking, falling, getting up — retrains its tiny models on the board, checks them, and keeps a new model only if it is actually better. It wakes up better than it went to sleep. A person can teach it too: show it a pose or a photo, and it learns that as well. Tonight it ran live at this hackathon: dozens of autonomous dream cycles, a pose taught by a human and relearned on the board in ~25 seconds, a local Qwen3 voice that answers from the robot's own real state, and a camera that learned to recognize the robot under test from a handful of photos. No cloud anywhere. One system, many bodies: today it is tested on the Hiwonder PuppyPi, and the same engine runs on a $1 walker body. The intelligence lives inside the robot — where it should be.
Built at ExecuTorch Hackathon — IoT Track
REM PhysicalAI
A robot that dreams itself better — a self-learning system for the Arduino Ventuno Q. Robots leave the factory and never learn again. REM PhysicalAI changes that. REM PhysicalAI is a self-improvement system that installs on a robot. Give a robot a way to act, a camera to see, and a computer to think — here, an Arduino Ventuno Q running ExecuTorch, 100% on-device — and it learns by itself: it commands an action, watches what happened, and relates the two. When nobody is using it, it dreams. In its dream it practices in simulation — standing, walking, falling, getting up — retrains its tiny models on the board, checks them, and keeps a new model only if it is actually better. It wakes up better than it went to sleep. A person can teach it too: show it a pose or a photo, and it learns that as well. Tonight it ran live at this hackathon: dozens of autonomous dream cycles, a pose taught by a human and relearned on the board in ~25 seconds, a local Qwen3 voice that answers from the robot's own real state, and a camera that learned to recognize the robot under test from a handful of photos. No cloud anywhere. One system, many bodies: today it is tested on the Hiwonder PuppyPi, and the same engine runs on a $1 walker body. The intelligence lives inside the robot — where it should be.
Keep exploring what builders shipped.
Magic Mirror
Magic Mirror
Project brief ExecuTorch Hackathon — Founders, Inc., October 17–18, 2026. Everything below runs on the Arduino VENTUNO Q on the desk: no cloud, no account, no network needed once the board is flashed. The pitch A store fixture that puts the shopper in the product photo. Stand in front of a camera and you appear wearing the garment — the head cutout, the body matte, the garment placement and the generated try-on image are all computed on the board's Hexagon NPU, with the web page that shows the result served from the same device. What's built and measured Piece Where it runs Cost Virtual billboard (BiSeNet head matte → product photo) NPU 12 ms compute, ~1 fps end to end Person segmentation (SegFormer human parsing) NPU 24.9 ms Generated try-on (Stable Diffusion + ControlNet-Canny) NPU ~0.3 s per step Face swap into a studio shot (GHOST, Apache-2.0) NPU 82 ms, fp16 LLM text generation (Gemma 4 E2B) NPU 188 tok/s prompt, 16.3 tok/s gen, 494 MiB Vision Q&A (InternVL3-1B) NPU 0.5 s encode, ~33 tok/s Live speech (Whisper-tiny → Gemma translate) NPU 1.8–2.6 s transcript, ~9 s translated LLM baseline (Llama 3.2 1B, XNNPACK) CPU 10.1 tok/s The NPU numbers are the point: prompt eval is ~12× the CPU baseline on a model twice the size, with resident memory a fraction of it, because the weights live in the NPU context. How it fits together Hugging Face model → optimum-cli export executorch → quantized .pte → ExecuTorch runtime. The same .pte runs on the Mac and on the board, which is what makes iterating on a laptop and demoing on hardware the same workflow. NPU models go through the track bundle's QNN AOT path (Linux only) and are driven by a warm persistent runner — 468 ms per inference launched cold, 26 ms held open with scratch on tmpfs. Two independent routes to the same effect, kept separate on purpose: overlay moves a garment onto a live frame (garment_overlay.py, garment_clip.py), and face swap moves the shopper's face into a studio shot that already has the garment, lighting and pose (board_faceswap_*). Hardware QCS8275, Cortex-A78C big cores, 14 GB RAM, Ubuntu 24.04 aarch64 over adb — not Android, despite answering adb, and considerably better than its published spec. No root; the compiler, the runners and the Python environment were all installed without it. Status and limits Board path: working, demoed, including the web UI and the walkthrough deck. Android/XR app (android/): compiles, never run — no device was attached. macOS cannot do QNN ahead-of-time compilation, so NPU work happens on the board. Thin segmentation classes (arms) lose accuracy at 8a8w; silhouettes don't. The board reboots under heavy load, so long board-side jobs are resumable.
Harshith Ande's team
Kamikaze
Kamikaze is an edge-AI vision system built on the Arduino VENTUNO Q that converts physical consumer items into interactive 3D digital twins in real time. Powered by an on-device Qualcomm Hexagon NPU, the pipeline uses YOLO-World, DINOv2, RapidOCR, and HSV color gating to identify smartphones and beverage cans during a guided five-second scan. An integrated FastMCP server then resolves metric-accurate 3D models by cross-referencing local CAD caches, 1:1 dummy repositories, and procedural mesh generators with live-stitched label textures—all while enforcing a zero-cloud privacy boundary. The resulting Smart Kiosk interface pairs Modulino RGB LED feedback and real-time hardware telemetry with an auto-orienting 3D viewport that dynamically mirrors the physical pose of the object in the user's hand.
Team AAA
Depth Perception
Collating Arduino distance sensor readings with video to train camera-alone depth perception