Producing hardware and software tools for controlled-environment agriculture production, Motorleaf leads in the development of artificial-intelligence software that enable growers to better monitor and control production and predict crop yields. Motorleaf’s agronomist.ai platform provides a virtual agronomist assistant that guides growers in adjusting growing conditions so they can replicate grow protocols with precision.
We develop software tools and machine learning algorithms to assess growing conditions and predict crop yields in indoor grow operations--most importantly, commercial greenhouses for vegetable cultivation.Greenhouse producers face many challenges in predicting their weekly yield in produce; the manual process is imprecise and often compromised by human error. Yields fluctuate depending on indoor growing conditions (e.g., humidity) and outdoor conditions (e.g., amount of sunny days).
With artificial intelligence, we can better predict yield in greenhouses. Using data from sensors (e.g., light levels), environmental controls (e.g., temperature) and weather information, we train machine learning algorithms to monitor growing conditions and then associate that information with yield. We also use machine vision to assess the rate of growth of foliage of plants, which is also associated with yield. We develop algorithms for each greenhouse. Once trained, our software tools can predict yield better than current methods.The results from our client SunSelect, a large-scale greenhouse producer of tomatoes in California, are exemplary. Using our software, they were able to reduce errors in predicting tomato yield by 72%. Our SunSelect client estimates that our AI-powered yield prediction tool can save them $70 000 a week in costs associated with otherwise imprecise yield calculations. From these results, SunSelect has abandoned calculating yield predictions manually--Motorleaf’s machine learning algorithms now perform this task.