Removing barriers to the Agri-food Data Revolution
Big data, agri-tech, artificial intelligence (AI), machine learning (ML), and internet of things (IOT) technologies have the potential to revolutionise productivity and sustainability in the food and farming sector. However, data-sharing is a pre-requisite. This report outlines some of the key barriers to effective data-sharing in the agri-food sector and proposes recommendations for overcoming them
Read the full report
Complete the form below and receive full access
Sector*:
You agree with our terms and conditions
Based on a report produced for the Food and Drink Sector Council's Agricultural Productivity Working Group; presented to Michael Gove, then Secretary of State for Environment, Food and Rural Affairs
Data has revolutionised industries across the board. In manufacturing, maintenance costs have been reduced by 40%, waste by 20% and yields improved by 50%. In pharmaceuticals, big data is optimising innovation – McKinsey's Global Institute estimates that these changes are worth $100bn annually. And amongst sales and marketing functions, customer retention is 300% higher in organisations who use data effectively.

Despite these advancements in other industries, it is the food and farming sector which has the greatest need for innovation. We are being asked to produce more food at a lower cost, on the same amount of land in a more sustainable way; a challenge which some commentators have labelled impossible. What's more, access to yield-enhancing pesticides is being curtailed. The political landscape regarding direct payments to farmers is in doubt. The livestock sector is facing public backlash and the commercial climate is bleak

The revolution is happening (almost)
Yet there is reason to be optimistic. Barfoots, the global food and farming company, is employing predictive modelling to streamline international supply chains. They hope that this will reduce waste and drive up their bottom line. BASF has combined multiple data streams and statistical calculations into a simple mobile application for growers. Its goal is to reduce pesticide run-off into watercourses by advising farmers when it's safe to spray. And Agrimetrics will soon release a suite of crop analytics derived from satellite imagery, which can be used to remotely assess irrigation and nitrogen requirements, plant health, crop lodging and yield predictions.
However, the greatest benefits are yet to be realised. In the future, real-time disease models will overhaul the way we apply pesticides. Predictive maintenance will eliminate all but the most unpredictable machinery breakdowns – saving farmers time and money. And a Data Marketplace will enable farms and businesses to create new revenue streams through the safe and fair sharing of their data.

Data is the catalyst for these innovations; data-sharing is a pre-requisite.

Barriers to revolution
In Spring 2019, Agrimetrics submitted a report to the Food and Drink Sector Council's Agricultural Productivity Working Group, which was presented to Michael Gove, then Secretary of State for Environment, Food and Rural Affairs. The report listed a range of cultural, structural and political challenges which are preventing wide-spread data-sharing in the agri-food sector, and recommendations for overcoming them. A modified version of that report is available here.

I don't trust you
The report lists 12 challenges, including a widespread lack of trust. Trust is foundational for data sharing. We must have faith in organisations to use our data responsibly and within pre-defined limits. This is especially true for businesses, where propriety data is often viewed as commercially sensitive and a competitive advantage.

Unfortunately, perceived poor behaviour amongst a minority of corporate actors and a widespread mistrust of the motivations behind the gathering of data by the public and private sectors has created an understandable cynicism.

This has not been helped by a slowness in the sector to adopt independent data governance accreditations, such as ISO27001. These accreditations are common in other data-rich sectors such as Finance and Healthcare. Typically, they will involve regular audits by independent accrediting bodies, who assess the rigour of data governance and security practices against best practice and a recognised framework.

Yet in a 2019 survey by the firm Supply Intelligence, just 2 of 36 agri-tech organisations were able to confirm ISO27001 accreditation (Agrimetrics and Shearwell Data).

Agribusinesses with ISO27001
ISO27001 is an independent data governance accreditation. This survey was conducted by Supply Intelligence in April 2019, who contacted 36 leading agri-tech organisations
6%
ISO27001 accredited
25%
Aware of ISO27001, without accreditation
69%
Not aware of ISO27001
Farm data is not digital data
As many as 60% of UK farms are still not keeping digital records, with many opting for paper-based record keeping. Of those who have started to digitise, records are often held in desktop-based, legacy IT systems. This presents obvious problems. Even if farmers were willing to share their data, that data is rarely in a shareable format. Furthermore, the means for data-sharing often do not exist.

Positive efforts are being made. Gatekeeper – a crop management software provider whose users manage 60% of the UK's arable land – have begun the progress of migrating their desktop products to the cloud. This will enable data to be aggregated, analysed and shared more easily. Muddy Boots – a competitor – now only offer cloud-based software. And a raft of new device manufacturers and application developers have built data-centricity into their business models.

Of course, farmers still need to agree to their on-farm data being used, which requires trust.

What's more, data must be interoperable. The predictive model developed by Barfoots, referenced earlier in this article, required combining and analysing multiple feeds of data, which were not automatically compatible.

This lack of compatibility can be overcome; however, it is more cost-effective to collect the data in the right format. This would require A Code of Conduct for Data Sharing and Common Standards for Interoperability – two key recommendations of the Report...


Read the full report for free