WeatherXM is a DePIN og, not exactly in the "consumer app" category.
We manufacture crypto enabled weather stations (hardware) and operate global weather infrastructure. We do sell consumer grade hardware but we target enterprises with the data & services.
Raised ~ $12m, Greek founders.
100%. In @openclaw we replaced generic eval rubrics with a simplified Elephant-Goldfish review loop: main agent (Gemini Argon) holds the full product/customer context ("Elephant") and stress-tests cold subagents (Claude CLI Fable) to strip out over-engineered AI slop before it hits production.
https://t.co/wCJjwESxEY
Letting coding agents iterate on physical ESP32 screens is a great move, but the ultimate unlock is using the ESP32 as a tunnel for agent to reach other LAN devices and then have agents discover each other 😈
I just did an experiment with Agent SDK from Meta Muse, on our weather station gateway (ESP32) and the use-cases we unlock are wild! e.g. think clawhub again but this time agents use their sensors to do weather nowcast instead of slop gossip.
https://t.co/LmM0mMqrUk
Digging into esp-openclaw-node now, looks better fit than Meta's to merge in our default firmware. So instead of a bare indoor weather screen, the OpenClaw gateway (and HomeAssistant) gets:
1. A 4" 480×480 RGB canvas + buzzer it can push custom cards/dashboards
2. Real rooftop ground truth (wind gusts, rain rate, UV, solar, pressure) + Sub-GHz LoRa radio to sniff RF stuff, like Meshtastic traffic, FLARM / OGN / FANET / Open Drone ID, etc
...and... for anyone running @openclaw in the cloud:
3. because our gateway is on the home Wi-Fi, adding somekind of https://t.co/bVC4B98GH8 + lan.fetch node commands (see our Muse implementation) turns the weather console into a zero-VPN smart-home bridge, so the cloud OpenClaw agent can pair live rooftop wind/rain with local Home Assistant, Shelly, HVAC or solar inverters on the LAN.
Most AI agents can tell you the forecast. This one looks at the sky outside my window and measures the actual weather on my roof.
We ported @Meta’s open-source Muse Gadget SDK to the @WeatherXM D1 (WG1200), turning a weather station into a Physical AI test bench:
📡 Live rooftop Sub-GHz LoRa sensors (https://t.co/KLuSxbHTYb)
👁️ Eyes on the sky via local LAN IP cameras (lan.fetch)
🎨 A 480×480 generative canvas (display.draw_file)
🔄 Reversible Dual-Boot (stock WeatherXM OS stays untouched)
Open-source repo & 1-click Web Flasher 🧵👇
We could add an fpga in a traditional motherboard and have AI reprogram it when needed but in my pov, todays software abstraction is useful and enough for agents to do miracles already.
Look at this idea from Meta's Muse Agent transforming microcontrollers to smart home trojan horses
https://t.co/LmM0mMqrUk
Most AI agents can tell you the forecast. This one looks at the sky outside my window and measures the actual weather on my roof.
We ported @Meta’s open-source Muse Gadget SDK to the @WeatherXM D1 (WG1200), turning a weather station into a Physical AI test bench:
📡 Live rooftop Sub-GHz LoRa sensors (https://t.co/KLuSxbHTYb)
👁️ Eyes on the sky via local LAN IP cameras (lan.fetch)
🎨 A 480×480 generative canvas (display.draw_file)
🔄 Reversible Dual-Boot (stock WeatherXM OS stays untouched)
Open-source repo & 1-click Web Flasher 🧵👇
I have also updated your web flasher with our hardware but wanted a bunch of extra features specific to a safe dual boot with signed firmware, so decided to spin a new flasher that also does the BLE provisioning (replacing the mobile app).
Overall this type of AI hardware integration opens up a great user experience, eg my default prompt is for Muse to fetch multiple forecasts, find the best fit for my actual weather conditions, thus likely the most accurate forrcast, and draw a weather card with a random style on the device. That's great fun but the real impact would be to allow agents to discover and collaborate on solving problems together, like synchronized sky photos for nowcast. Looking fwd to your next steps in this direction :-)
hey @anantn quick heads-up on two things I've put together for Muse gadget ESP32:
Web-based Flasher & BLE Provisioner (https://t.co/c6k1jh5zMi):
To make it frictionless for the thousands of WeatherXM station owners to try Muse alongside their stock weather station firmware. I've built a browser-based USB flasher + Web Bluetooth provisioning flow (with a small Worker proxy for token minting/VM lookup). The BLE/token setup is admittedly a bit hacky right now — hope you don't mind us shipping it for our community testing!
LAN.fetch on ESP32 (PR #299):
We also opened PR #299 implementing lan.fetch on the ESP32 so the gateway can pull snapshots from local LAN sky cameras (and local HTTP endpoints) and forward the images to Muse for live visual sky/weather analysis. I hope this was within your plans for ESP32.
Would love your thoughts when you get a chance to take a look! The AI + hyper local weather usecases are endless and amazing!
@1000ouz Muse firmware flashing and BLE provisioning merged
on https://t.co/GgqwRdNGLr so the entire process happens via web broswer, no need to install anything on local machine.
Most AI agents can tell you the forecast. This one looks at the sky outside my window and measures the actual weather on my roof.
We ported @Meta’s open-source Muse Gadget SDK to the @WeatherXM D1 (WG1200), turning a weather station into a Physical AI test bench:
📡 Live rooftop Sub-GHz LoRa sensors (https://t.co/KLuSxbHTYb)
👁️ Eyes on the sky via local LAN IP cameras (lan.fetch)
🎨 A 480×480 generative canvas (display.draw_file)
🔄 Reversible Dual-Boot (stock WeatherXM OS stays untouched)
Open-source repo & 1-click Web Flasher 🧵👇
wg1200 has 3 firmware slots, so you should be good to play around with https://t.co/95fO59xRQN without even loosing your wifi settings. The Muse firmware is more risky though, was only developed this.. weekend so there is more risk involved. I'm trying to merge it in our flasher and make it safer ...will update repo and post something new in a while
@1000ouz eh.. yea.. I guess I wanted an excuse to make a better muse gadget web flasher that also helps with provisioning which is also messy after the flash.
brb
Depends on distance of cameras and wind speed but ~100msec sync is enough. According to research a 0.1sec skew in strong wind (15 m/s): cloud moves 1.5 m → ±3 m ceiling error (0.3%). A 1 sec skew would make the error 3%
So ESP32 NTP is all we need, just need to make sure snapshot is fresh from the camera
We ported Muse to the @WeatherXM D1 rooftop LoRa weather station, and upstreamed lan.fetch in PR #299 so any ESP32 Muse gadget can pull local LAN sky-camera snapshots.
Single-station Physical AI works today, with cryptographic proofs on data provenance in our case.
The real unlock is an agentic "SETI@home" for weather nowcasting: neighboring Muse agents discovering each other, tasking synchronized sky captures, and grading forecasts against rooftop weather readings.
@alexandr_wang when will the Muse agent-to-agent substrate be ready so we can change the face of meteorology forever?
@1000ouz Hey, thanks for giving it a spin.
Is https://t.co/FzRZ8QaSC3 working for you? That's a better designed flasher, the Muse one was quick and dirty patch of what was in their repo.
6/ Want to experiment with Physical AI on real outdoor instruments now!? We ship globally, like our network !
Use code MUSE50 for $50 off per station on the WeatherXM D1 Wi-Fi ($199 → $149 Network Edition)!
🛒 https://t.co/DmbgwB4YEI
Get two or more to test multi-agent coordination today!
And subscribe for off-grid LoRa Mesh multi-hop that is comming next:
https://t.co/NQhgWiuDPV
5/ Where this goes next: Self-Assembling Physical AI Nowcasting 🌍
1️⃣ Discovery & Mesh Rendezvous: Stations within a region, each run by its own independent Muse agent, find each other via a signed rendezvous and advertise what they can observe: camera bearing, live sensors, scored history. No central planner.
2️⃣ Peer-to-Peer Tasking: "Capture the sky at T+90s, signed—I'll do the same for you." Two synchronized sky views + wind vectors → stereo cloud height, optical flow, and a shadow-arrival countdown drawn on the neighboring rooftop's D1 screen. Agentic nowcast!
3️⃣ Graded by Nature: Every prediction is scored against physical ground truth—did the rooftop solar irradiance (W/m²) actually drop at the predicted minute? "Insufficient geometry" is a valid answer, and error bars get published next to the wins. Agents get better in every turn.
📡An agentic SETI@home for weather nowcasting! Instead of just donating idle CPU cycles, Muse AI agents contribute live rooftop telemetry, night-star auto-calibrated sky-camera views, and peer-to-peer coordination.