Turning data into decisions

Visualising the opportunity that data holds for the agrifood sector is often hard, but a new paper recently published by Wiley in the Journal of the Science of Food and Agriculture (JSFA) gives five good reasons on how bringing complex data together can drive more informed decision making.

‍

Agrimetrics was one of eight co-authors of the paper – How should we turn data into decision in AgriFood. The paper, published by Wiley’s Joint Science and Farming Agriculture, was the result of a workshop run by the Society of Chemistry and Knowledge Transfer Network.

The group identified a number of challenges to unlocking the opportunities held in data such as ‘low inter‐operability of different data sets, silo mentality, low willingness to share data and a significant skills gap.’

The paper also singled out five major challenges to benefit from data connectivity and analysis:

1. Improving livestock production

Livestock production systems vary enormously. There is data available, however, that can provide much more insight on farming management and systems to deliver meat at optimum specification and therefore best business returns. Some of the parameters the group identified that are fundamental to driving greater meat production efficiency include all these stages in the food chain:

• pre-birth (fertility, gestation length, birth rate)
• rearing (growth rate, feed conversion ratio, disease resistance, mortality)
• finishing (weight, yield, fat class); slaughter (abattoir process)
• processing (butchery, processed meat products)
• retail (meat colour, fat content, pack transparency and traceability).

2. Automation and robots

A shortage of labour will require automation and robots, concluded the paper, to address the urgent need to produce more food at higher environmental and quality standards. Data at key points in the food production system can identify where most value is created and reduce inefficiencies.

Furthermore, digital platforms help to bring buyers and sellers closer together, enabling vital information to flow up and down the supply chain. The other major value of robots is automating the collection and accuracy of data either in the field or in food manufacturing.

3. Food provenance and safety

The authors describe how data plays a key role in consumer trust and safety – helping to establish the provenance of products and the conditions under which they have been brought to market. Data can help to develop real-time prediction of emerging risks to food safety and fraud such as the horsemeat scandal in 2013.

They describe how commodity prices, consumer price index, exchange rates, extreme weather, pest and disease incidents, changes in regulation and standards, profit margins and production capacities can all be used to develop early warning systems for food fraud. Deploying algorithms based on machine learning and statistical methods that aggregate all layers of such data and detect anomalies can collectively highlight potential issues.

4. Shorter supply chains

Digital technologies could lead to on-line trading platforms, which may also help to open-up the food market and bypass some of today’s distribution channels. The paper concluded that greater understanding by farmers about consumer behaviours can help to guide business planning, marketing and new products.

5. Managing risks and uncertainties

Farm businesses succeed or fail on productivity and prices, which will become increasingly important as subsidies move to paying farmers for environmental measures. Analysing data to create algorithms on precise correlations between each commodity over time will help to accurately forecast future price – a critical parameter in determining a farm’s profitability.

How should we turn into data into decision in AgriFood?

Published by Wiley and jointly authored by Serazetdinova, Liliya; Knowledge Transfer Network, AgriFood Garratt, James; Enviresearch Ltd, HQ Baylis, Alan; Nuvistix Innovation, HQ Stergiadis, Sokratis; University of Reading, School of Agriculture, Policy and Development Collison, Martin; Collison and Associates Limited, Davis, Simon; Agrimetrics Ltd.

‍