Microsoft: “Great potential for AI in agriculture”

Ms Roessler is Director Industry Solutions for Chemical & Agriculture at Microsoft Corporation and an expert in the field on how a software company like Microsoft can be a platform for new technologies and information streams that are entering the global agricultural sector at a fast pace. We caught up with her in the run-up to the second Global Future Farming Summit on 6-7 November 2018, organised by Proagrica in cooperation with Wageningen University and Research Centre, this November.

In agriculture, do you think are we still too much focused on (only) gathering the data, instead of the analytics?

“For starters, it is difficult to make a global comparison, because practices and maturity of big data in agriculture vary significantly. This is also because the criteria for how data is collected and regulated by individual markets is still being determined. Adding to this complexity is the fact many agtech startups create their own systems to aggregate data. So there are challenges on both the data gathering (connectivity, costs, scale) and in the analytics. Regarding the latter, much more can be done with putting different data sources together. This is for example what we aim for in our project FarmBeats. Here we aggregate field-level sensor data, together with drone imagery/video data or satellite imagery data to predict the conditions on a farm”.

What is the reason that Microsoft is so engaged in agriculture and what are some of the projects you are involved in?

“Some time ago, we noticed that big agricultural companies were trying to make sense of things like big data, artificial intelligence (AI), machine learning and analytics. These things rapidly moved their way into agriculture and most companies were not equipped to handle or deal with these new technologies. The need for technology companies like Microsoft to help them navigate this shift is critical, so they pro-actively came to us for advice and help. I want to emphasise that Microsoft doesn’t want to become an agricultural company, but we want to be a fundamental platform for agriculture data, analytics and technological advancements in agriculture, which in turn help farmers and agricultural companies improve productivity and address sustainability issues. We do this commercially with the products we offer to farmers, but also on a more academic level, where we want to create a community around topics like AI and machine learning in agriculture. This is reflected in our programme ‘AI for Earth’, where we work with non-profit and universities globally to advance the development of agriculture related models and API leveraging machine learning and AI to address challenges of agriculture, sustainability, bio-diversity and climate and show the “art of possible”. With our partners, such as The Yield we are working on precision farming to help farmers make faster decisions in managing weather and environmental factors; providing on-farm sensors and customised information services to help farmers ensuring better productivity and profitability.”

Why are you so enthusiastic about AI and machine learning in agriculture?

“It offers so many possibilities. For starters, AI and machine learning will allow us to collect better information, precise input and provide proactive recommendations at scale. It’s going to allow us to make more informed decisions on the farm and along the food value chain, helping agriculture customers predict demand, be able to simulate certain conditions, warn us for yield damaging events and proactively help customers make recommendations for improvements based on sets of aggregated data.

There is a strong momentum of AI in automation and robotics, such as for autonomous working equipment on farms, intelligent robotics, precision spraying etc. What I find most exciting is that AI will allow us to identify patterns in nature that we would likely never been able to see with a human eye. It can also be done much faster than before. AI can answer questions like: How does plant genomics, environment, humans, animals correlate and interfere? How is it impacting plant growth and health? This in turn will lead to making better decisions and being more proactive as a farmer. And this is really exciting stuff.”

Sounds exciting indeed, but what is needed to make the next steps?

“If we truly want to have a global impact on sustainability and food production and safety, we’ll have to think about data standards, data sharing, benchmarking and analytics on aggregated sets of data. Today, we see farmers who are reluctant to share data and agtech start-ups who have built closed systems, which will keep things from evolving. In addition, there is fantastic innovation happening at universities, that is not often known or accessible. Many large agricultural companies are working in data silos instead of data lakes. The industry has to converge as an ecosystem, where all the players can remain independently, but can share data in a secure environment, run analytics on complex systems of combined sets of data or provide a secure space where IP from one participant can come together with data from another, without losing ownership or control of either. This requires trust and our company is helping to build those secure, trusted data platforms providing the foundation of multiparty-data sharing and consortia to tap into collective insight, awareness and action”.

You are one of the speakers at the upcoming Global Future Farming Summit. Why are summits like this, focusing on technological changes in agriculture, so important?

“First of all, the Global Future Farming Summit is bringing together agriculture experts from the academic world, commercial companies, and technology firms. This mix of people creates a fantastic opportunity to learn from each other and to share practical examples from the field. It is about “learn it all” instead of “know it all”, which is also deeply rooted in Microsoft’s culture. We also see it as an opportunity to share examples of the deep impact technology can have on productive and sustainable agriculture. But, fundamentally those kinds of events are going to be the birthplace for those cross-industry collaboration initiatives. The “garage” where some of these secure, trusted data platforms are envisioned and planned. I therefore look very much forward to this fantastic event”.