Microsoft has been working on new ways for agritech workers to extract data, and it is hoping that open sourcing will lead to knowledge sharing. As part of its ongoing research into agritech, Microsoft is introducing FarmVibes.AI, a set of open-source AI models for farmers and others working in the industry. FarmVibes.AI aims to advance data-driven agriculture, according to Microsoft’s AI blog. The blog post outlined the case study of a US farmer and software engineer who has been using the tools to inform decisions he makes on his farm. FarmVibes.AI is the first release as part of a wider Farm Vibes project, according to Microsoft. The tools run on Microsoft Azure. The company is making FarmVibes.AI open source to encourage data scientists and agritech experts to leverage it in their own research so farmers can eventually benefit from data gathering tools – no matter how rudimental their tech skills are. FarmVibes.AI is available to download from GitHub. The research team at Microsoft is currently working on more tools that will be made open source in the future as part of Farm Vibes. These include FarmVibes.Connect, a tool designed to improve internet connectivity in rural places. The current open source offering, FarmVibes.AI, offers several main components. Its ‘Async Fusion’ tool combines drone and satellite imagery with data from ground-based sensors to offer insights on planting and crop fertilisation. The ‘SpaceEye’ tool removes clouds from satellite imagery using AI. The ‘DeepMC’ algorithm uses sensor data and weather station forecasts to predict weather conditions. There is also an analytics tool that estimates how various farming practices would affect the amount of carbon sequestered in the soil.
