What Is MAIN? A Plain-Language Guide

MAIN stands for Microalgae Artificial Intelligence Network, but the plain-language version is simpler: it is a small computer that grows spirulina for you and keeps an eye on it around the clock. Spirulina is a microscopic, spiral-shaped cyanobacterium (a photosynthetic microbe) that lives in warm, very alkaline water. Keeping it healthy means holding several conditions in balance at once, and those conditions drift all day as the culture drinks in light, feeds, and grows. MAIN's job is to notice that drift early and correct it gently, so a living culture is never left to guesswork.

MAIN senses its culture through six simple readings, taken about once a minute. It measures pH (how alkaline the water is), temperature, and light. It watches density, meaning how thick and green the culture has grown, which is really a measure of how much spirulina is now in the water. And it tracks dissolved oxygen (the oxygen the culture gives off as it photosynthesizes) and TDS, the total dissolved salts and nutrients. An optional tiny camera can add a microscope-style view to catch contamination and help estimate density. None of these numbers mean much on their own. MAIN's value is in reading them together, the way an experienced grower can glance at a jar and simply know something is off.

When a reading drifts, MAIN can respond with a few careful actions. It can dose bicarbonate, which is the culture's carbon source and its natural pH buffer rolled into one. That one matters because when the pH climbs too high, it usually means the culture has eaten through its available carbon and needs a top-up. MAIN can also add liquid nutrient, run an air pump to stir the culture and let built-up oxygen escape, switch on a heater, or adjust the light. The moves are deliberately small. MAIN nudges the culture back toward comfortable; it does not yank it.

Before it makes any move, MAIN thinks ahead using what engineers call a digital twin, a physics-based simulation that behaves like a virtual copy of your culture. It draws on the well-understood biology of how spirulina responds to light, warmth, alkalinity, and oxygen to forecast what the next day of growth should look like, and whether a proposed action would genuinely help. This is where some of spirulina's quirks show up. More light, for instance, is not always better: past a certain brightness the culture is harmed rather than helped, so MAIN aims for a productive middle band instead of just turning the lights up.

The most important thing to understand about MAIN is that the AI only proposes; it never acts unchecked. Every suggested action passes through two safety layers: a set of fixed hard limits that no action may cross, and a simulation check that blocks any move predicted to stall growth or push the water out of its safe range. The system fails closed, which means when it is unsure, it does nothing rather than gamble with the culture. Beneath the software sit independent backstops, including limits built into the low-level firmware and a physical emergency-stop button you can press by hand. And if the internet drops, a simpler set of rules keeps running on the device itself, so the culture is never left unattended.

MAIN reactors can also learn from one another through a network called PhycoNet. When one reactor works out a better way to keep a particular strain happy, it shares only the lessons, meaning the tuned settings its simulation has learned, and never your raw sensor logs or camera images. Any settings arriving from the network are checked for sensible values before they are ever allowed to touch a live culture. The idea is that the whole network gets steadily better at growing spirulina, while each grower's own data stays their own.

A note on what MAIN is and is not. It is an experimental, open-source, educational project, built and shared in the belief that science should be open to everyone, and it is run as a non-commercial effort by a young creator. It is a tool for learning how a living culture behaves and how careful automation can look after it. It is not a health or medical product, and it makes no promises about what spirulina can do for you. If you want to follow along or ask questions, the project lives on YouTube at @OrrBiologicals, and you can reach the team at service@orrbiologicals.com.