In Which Data Universe Do You Need to Play With Artificial Intelligence?

In today’s world, farmers themselves optimise their operations. In order to do so, they gather information from three data universes. These three universes are:

1. External data

For example, weather data, market prices of agricultural products and inputs, and developments in technology offerings. Typically, farmers source this data from, among others, the trade magazines, their suppliers and off-takers, the internet, and trade fairs.

2. Internal operational data

For example, yield data, feed conversion ratios as applied data for fertiliser, crop protection chemicals and animal health products, energy costs, labour productivity, etc. This data is available in the farmer’s management information system.

3. Financial data

Such as cost price calculations, liquidity situation, fiscal reporting, and investment decision evaluations. This data is compiled by accountancy firms, banks, and farmers themselves.

Farmers use the data from these three universes for operational, tactical, and strategic decision-making. Typically, financial data plays a strong role in strategic decision-making (e.g. long-term investment decisions), but a much smaller role in tactical decision-making and barely a role in operational decision-making. For the most part, operational decision-making is about straightforward agronomic optimisation. We expect this to change. Operational decision-making will increasingly include weighing the costs and benefits of agronomic measures. For example, a fertiliser application will not only depend on the yield a farmer wants to achieve and the availability of nutrients in the soil, but also on the cost of the fertiliser and the expected value of the to-be harvested crop.

As described, agronomic optimisation and financial optimisation increasingly go hand in hand. Operational agronomic decisions will increasingly be evaluated as if they were investment decisions, by focusing on the return on ‘investment’. The complexity of optimisation of agricultural practices increases, because it has to take into account agronomic and economic variables and needs to be tailored to the specific situation at the farm, field, or within the site-specific situation. This requires linking data from the three universes in order to establish what the optimal site-specific yield is and how to achieve it.

For farm inputs companies that want to support farmers’ decision-making processes with artificial intelligence, this intertwining of data universes has substantial consequences. In order to offer true added value to farmers, companies will need access to data in all three universes—and they need to be able to tackle the complexity along three dimensions, prices, agronomic processes, and variability in circumstances. Therefore, the focus for these companies will not only have to be on creating access to farm data, but also to external data, in order to develop and apply their artificial intelligence.