Why.
Ask a simple question about a small piece of the natural world — is this creek warming? when did the oxygen drop? what did last summer actually do to this meadow? — and the honest answer is often that nobody knows. Not because nobody cares, but because nobody was measuring.
Continuous environmental monitoring exists, and it works. It's also ludicrously expensive. A single commercial telemetry station runs thousands of dollars before the cellular or satellite subscription starts, and the data often lands in a platform you rent rather than own. This has prevented many of those closest to the land – watershed groups, small farms, tribes, researchers, and activists – from being able to access continuous monitoring. Mycelium is designed to change that.
Growing up in west Berkeley, one of my favorite places was the short, daylighted section of Codornices Creek on the Ohlone Greenway near Gilman Street. This segment of creek is far from wild, it sits under the BART tracks and flows into concrete tunnels on either side. But on warm days after preschool my godfather would take me there and we would "fish" with a piece of string tied to a stick, a bent nail, and a piece of cheese. I vividly remember the trout nibbling at the cheese while crayfish crawled along the creek bottom and hid in the undercut banks. I (fortunately) never managed to catch a fish, but this piece of urban nature was magical and precious to me.
Today that same section of creek is far more desolate, floating plastic bags have replaced the trout, which have long since disappeared. You can still see the occasional crayfish crawling the bottom, but they look more like survivors of a disaster than members of a functioning ecosystem. Over the intervening years this short section of creek has been degraded to a mere shadow of its previous self, a process which is repeating the world over in uncountable small fragments of the natural world.
Despite living nearby my whole life:
I could not tell you why this happened.
Was it warmer waters, a diesel spill, or something completely different that killed off the trout? To protect these precious remnants, I realized that we first needed to understand what was happening to them, but the existing tools are far too expensive and complex. No one was going to spend $5,000 on a telemetered data logger for this short stretch of urban stream.
If I wanted to help save these places, I needed to figure out how to build telemetry that was cheap, reliable, and easy to deploy. It was after being inspired by Mark Qvist's phenomenal Reticulum ↗ networking stack that the idea for the Mycelium network was born: a distributed, open, hardware-agnostic environmental telemetry mesh that would allow anyone to cheaply deploy remotely monitored sensors.
Rather than a cellular or satellite backhaul, Mycelium uses meshed 915 MHz LoRa radios. As soon as you power on a node, it finds and connects to other nearby nodes automatically. So long as one node in the chain is connected to a gateway with WiFi, all connected nodes can send encrypted data down the chain to the gateway, which then posts it to an online database. Each deployed node strengthens and extends the range of the existing mesh, regardless of who operates it, and because all traffic is end to end encrypted, teams need not trust other node or gateway operators, or even know who they are.
The chained mesh nature of the network also allows networks to penetrate into areas without cellular coverage where previously only expensive satellite telemetry was viable. For example, say you needed to take water temperature measurements along the East Bay's Wildcat Creek. A single WiFi gateway node at the Tilden visitor center would connect the entire chain to the internet, allowing a node miles away to upload its data and be monitored remotely at zero ongoing cost, regardless of the fact that much of the watershed lacks cell coverage.
In order to understand the state of our natural world, we need better data. Real-time data with higher spatial resolution so it's not only the few most important places which get monitored. Though it's still in beta, I hope the Mycelium network can be one of the pieces which helps to fill in this gap, empowering academics, government agencies, and citizen scientists to fill in the blank spots in our knowledge of what is happening to the world around us.
Mycelium is young. The network itself runs end to end on the bench today, with its reliability layer tested by scripted failure drills against the live system. Field hardware (solar power, enclosures, real-world radio range) is still being proven, and I won't publish performance claims until it has been. If you are an engineer, researcher, or just someone with an idea you think could make mycelium better, I'd love to hear from you. Shoot me an email or issue a PR if you're interested in contributing.
The story of what happens next is partly yours to write.