With the world’s population growing rapidly, it will become more difficult to produce food quickly and sustainably in the amounts needed. By 2050, we’ll need to increase our food production by at least 50 percent.

Greenhouses and indoor farming provide a solution to produce fruits and vegetables faster, more safely, and with much fewer land and water requirements. However, it is difficult to find enough skilled labor for greenhouse production. Automation is an answer, but there is a lot still unknown about the greenhouse environment, about plants, and even about farming.

The Sonoma project aims to develop an autonomous system that can operate indoor farms efficiently and reliably. The concept also enables experts to remotely control production facilities, worldwide, so a fewer number of experts can run more facilities.


Our AI Approach

Sonoma combines both existing domain knowledge (such as crop modeling, greenhouse climate modelling, and controlled environment agriculture best practices) and modern machine learning algorithms, including Bayesian learning, deep learning, and reinforcement learning. The result will be an indoor farming solution that is both efficient and scalable.

Project Sonoma - Our Approach to AI



In the News

Microsoft Blog

Indoor vertical farming in Asia and beyond: Digging deep in data

When it comes to farming, Ken Tran digs deep – not in dirt, but in data. He doesn’t drive a tractor and he doesn’t pull a plow. But he is helping to sow the seeds of a new type of agriculture – one that is nurtured by machine learning and artificial intelligence. And, it promises to revolutionize how communities across Asia and the world feed themselves.

February 27, 2018
Asia News Center