Building Physical AI, in the open

Ubuntu and NVIDIA JetPack are an excellent development environment for a Jetson. WendyOS adds a reproducible, immutable base with atomic A/B updates for teams operating edge AI devices as a fleet.

RAUC and Mender put reliable OTA updates at the center of their products. WendyOS includes signed A/B OTA with automatic rollback as part of a broader Physical AI platform for container deploys, hardware access, device identity, and telemetry.

Docker runs one container on one machine you're logged into. Deploying and updating a fleet of robots needs OS updates, hardware access, device identity, and DDS discovery. Here's how WendyOS compares to a hand-rolled Docker + ROS 2 stack.

balena pioneered container fleets for IoT. WendyOS takes the same idea to robotics and Physical AI - GPU and camera access, ROS 2, a local build-and-stream loop that works offline, and Apache 2.0 licensing. Here's how they compare.

Which local LLMs to run on the NVIDIA Jetson Orin Nano 8GB and Jetson AGX Thor with the Wendy llm template, plus how to install, scaffold, and deploy them with the Wendy CLI.

Deploy Wendy applications to NVIDIA DGX Spark with full GPU access: install the Wendy CLI, install wendy-agent on the Spark, add the GPU entitlement to wendy.json, and run wendy run.

We are releasing a free NVIDIA Jetson Thor 3D model for renders, diagrams, product videos, robotics explainers, and Physical AI visualizations.

At the EUROPE EMBODIED Robotics Hackathon, a three-person team built an autonomous facility inspection stack for the Unitree Go2 quadruped - ROS 2, SLAM, Nav2, and LLM control - and deployed it on a Jetson Orin Nano booted with WendyOS.

The Wendy CLI now builds your robotics apps with Apple Container by default on Apple silicon Macs. Linux container builds run in lightweight native VMs instead of Docker Desktop, so wendy run and wendy build are faster with zero setup.

TeleCollect is a free app for iOS, iPadOS, visionOS, and macOS that records both video and per-frame skeletal joint data, giving you the raw pose datasets you need for your reinforcement learning journey with arms and humanoids.

We are releasing a free, high-quality NVIDIA Jetson Orin Nano 3D model for renders, diagrams, product videos, and robotics visualizations.

Walk through every file in the voice-ai-pipecat template, an always-listening voice assistant that runs faster-whisper for speech-to-text, Gemini 2.5 Flash for reasoning, and Piper for speech, all driven by a wake word and wrapped in a live audio visualizer.

Swift isn't just for iOS. With WendyOS you can scaffold a full-stack Swift app with a React frontend and a Hummingbird backend, then deploy it to a Raspberry Pi 5 or NVIDIA Jetson with one command. Here's why Swift is a serious language for robotics and edge AI.

Walk through every file in the RealSense template app - a Python FastAPI server that streams color, dual IR, and colorized depth from an Intel RealSense D415 over MJPEG, with a React control panel on top.

WendyOS brings a mobile-style developer workflow to the NVIDIA Jetson Orin Nano: connect over USB-C, deploy from your machine, and skip the monitor, keyboard, mouse, and adapter pile.

Deploy apps to your Raspberry Pi with a single command. Breakpoints, logging, and the full Wendy developer experience - now on the world's most popular single-board computer.

Wendy is now available as an early preview for developers, students, and tinkerers. Deploy containerized apps to Raspberry Pi, Jetson, or any Linux device with a single command.

Write firmware in Swift and deploy to ESP32 RISC-V MCUs using WebAssembly. GPIO, I2C, SPI, BLE, WiFi, and full OpenTelemetry observability - on a $5 chip.

We gave Claude Code one instruction and the wendy CLI documentation. It built a complete real-time object detection application with a React frontend, FastAPI backend, and deployed it to an NVIDIA Jetson - without a developer writing a single line of code.

We invited Claire Wang, an MIT Electrical Engineering and Robotics student, to run an experiment comparing the traditional NVIDIA Jetson Orin Nano setup process against WendyOS. The results were staggering.