Skip to content
Koidra
Book a demo

AI that acts on heat, mass and machinery

The agentic system for industrial processes

One platform that sees everything your operation produces, decides what needs to happen, and, where the physics allows, runs it.

Book a demo What autonomous operations actually takes

Proof

An AI ran a commercial crop better than expert growers. Twice.

Koidra won the international Autonomous Greenhouse Challenge twice, in 2018 and again in 2022. Each time the crop ran for months with nobody at the panel. That is the hardest version of the problem, and it is the reason we can use the word autonomous without an asterisk.

Aerial view of a commercial glasshouse complex, several ranges of glass with the packhouse and buffer tanks alongside
2×
Winner, Autonomous Greenhouse Challenge 2018 and 2022, a live crop each time
19.5%
Yield increase on cucumber against a control zone grown manually alongside it
17%
Energy savings in commercial greenhouse operations
20%
Dryer efficiency gain at Ayo Biomass, after a two-phase KoPilot deployment

Industries

Built for operations where output and energy are decided hour by hour

Manufacturing
Inputs vary load to load and every machine downstream absorbs it. The drift surfaces on a utility bill a month later rather than as an alarm.

Watched here: Energy per tonne

Greenhouse Farming
Yield is what the year turns on, and it still depends on which grower is on shift. Light and heat follow the same setpoints nobody has revisited since last season.

Watched here: Yield per square meter, and energy per kilo

Data Centers
Cooling is the largest controllable cost in the building, and setpoints are still tuned by hand against a comfort margin nobody has revisited.

Watched here: PUE and approach temperature

What our customers say

We are excited to partner with Koidra in this innovative trial. Their AI-based automation technology represents the future of sustainable agriculture. By integrating Koidra’s solutions, we aim to enhance our operational efficiency and continue providing high-quality produce to our customers.
Octavio Perez Rodriguez Director, Growing Operation at Nature Fresh Farms
Koidra’s dashboard visualizes the plant’s condition and real-time performance, enabling me to monitor it remotely. Thanks to up-to-date insights, I can lead cross-function teams, improve processes in a timely manner, and make data-driven business decisions.
Trang Do Continuous Improvement Manager, Ayo Biomass

The platform

Two ways in. One platform underneath.

Start with Maintenance Management, which needs nothing connected, or with DataPilot, which we connect for you. Neither needs the other. AI Advisory needs both and comes with both, and is where predictive maintenance comes from. KoPilot runs the process on top. Every product pays for itself on its own.

  1. Start here Maintenance Management It runs its own work
  2. Start here DataPilot It sees itself
  3. Comes with both AI Advisory It explains itself, says what to change, and runs predictive maintenance
  4. Autonomous control KoPilot It holds its own setpoints

What autonomous operations actually takes, or see how the four fit together.

One platform, because an agent can only act on what it sees

Koidra holds process sensors, energy meters, equipment history, maintenance work and the readings operators type on their phones in one place, against the same equipment and the same production batch. Most plants have this data. Almost none have it joined up, which is why almost none can automate anything above a single control loop.

Physics-aware AI, because a wrong setpoint is expensive

Our models are grounded in how heat, mass and machinery actually behave, rather than fitted to last quarter’s numbers. The model knows a dryer cannot heat faster than its mass allows, so it never proposes it.

That is what makes an autonomous decision safe enough to trust with a live plant, and it is where our research heritage sits.

The closed loop, and the envelope it obeys The process is read by a physics-aware model, whose proposed setpoints pass through an operating envelope that states the limits it may not cross, then reach the controllers, which write them back to the process. The loop runs every minute. The process dryer, boiler, hall, crop Physics-aware model knows what the equipment can and cannot do Operating envelope the limits it may not cross Controllers setpoints written back every minute, not every shift
The envelope is a separate box on purpose. It states what the loop may not do, and it is engineered for the site rather than learned from it, which is why a model that has never seen your plant is still safe to run on it.

How it works

One reading, four things that happen next

The platform treats a sensor reading and a number an operator types the same way. That is why an alarm can fire on a hand-entered moisture reading, and why a trend can open a work order before anyone notices it.

How a reading becomes an action A sensor reading and an operator's hand-entered reading both enter one operating record. From that record an alarm reaches a phone, a work order opens against the right asset, reports are written, and KoPilot adjusts setpoints. Process sensor every few seconds Operator's reading typed on a phone One operating record same equipment, same production batch, same clock Alarm on a phone Work order opens Report is written KoPilot adjusts setpoints both are the same shape to the platform
A historian never sees the operator's reading, and a maintenance system never sees the trend. Holding both makes everything to the right of the record possible.

See Koidra on your own plant data

Bring a week of readings from one line. We will show you what the platform sees, what it would have flagged, and what it would have scheduled.

Book a demo