Can on-farm technologies and “big data” support food and fibre product attributes that consumers value?
- What ingredients are in the food or beverage product (64%);
- How a company treats animals used in its products (44%); and
- From where a company sources its ingredients (43%).
We call these informational aspects of a product “credence attributes”, meaning that they give credence to our decision to purchase (or not purchase) a product or service, but can’t be directly assessed from the product itself, either before purchase (on the basis of colour or feel) or after purchase (on the basis of taste, for instance).
Characteristics such as “organic”, “environmentally responsible”, “grass-fed”, and “naturally raised” relate to the story behind a product. A product may communicate these through advertising, packaging, and other ways of telling the product story.
But consumers are also looking for authenticity and integrity in their food and other products. There’s a consumer backlash when the product story on the pack is in conflict with other data sources – such as claims in news articles or secret video footage.
We’ve been exploring ways that feeds of data from on-farm technology could be used to support the product provenance and credence story – or at least signal to farmers and their supply chain partners where checks and improvements should be considered. Here are a couple of examples.
Monitoring carbon footprint
Carbon life-cycle assessments (LCAs) are used to understand the extent to which production, manufacture, and distribution of a product impacts on climate change through deforestation or release of greenhouse gases such as carbon dioxide, methane, and nitrous oxide. We learn some interesting things from these, sometimes showing that shipping food products from the other side of the world can have a lower impact than growing products locally if the local environment is less hospitable.
Importantly, producing a Life-cycle assessment creates a model – a series of equations and if-then logic that describes the calculation. We can use this model with appropriate local farm and supply chain data to understand how management decisions and activities, timing and stock or crop productivity impact on emissions.
Automated systems on farms that capture data about crop production, livestock weights and production, and farm activities can also deliver data for a custom life-cycle assessment. Benchmark data across multiple farms and it becomes possible to identify the patterns of complete vs missing data, to understand how climatic constraints change emissions, or to identify outliers that need to be more closely examined.
A note of caution here: as we’ve learned from nutrient budgeting, farm systems can be varied and life-cycle assessment models are frequently based on the “typical”. An outlier result may indicate greater variation than the model can handle, rather than a more or less efficient farming system.
Demonstrating animal welfare
Animal welfare and the ability to live a healthy and natural life is another area of concern to consumers. Here too, metrics collected on-farm can be the subject of automated analysis to demonstrate good practices are followed.
In Europe where a premium is payable for “grass-fed” dairy in some regions, farmers are experimenting with the use of monitoring devices – smart tags and neck bands for example. These devices capture data that provide farmers with early warning of heats and potential animal health issues – raised temperatures, more or less movement, and reduced eating for example – but can also be analysed for patterns that only show up in outdoor grazing.
In other jurisdictions, veterinary product purchase, use, and reordering records can help to demonstrate compliance with animal health plans worked out between farmers and veterinarians, and hence demonstrate good welfare practices and appropriate use of medicines. Paper records have been used for this purpose for many years, but software technologies and automated data analysis can reduce the burden of data collection and the need for manual audits and analysis.
Some producers will find the thought of such automated systems invasive and potentially threatening. Certainly, given the potential for outliers, for good practices that just don’t quite fit the expected mould, and for technology glitch or human error, you couldn’t use these measures as legal baselines that determine “rights to farm”.
Nevertheless, application of technology and analytics such as these can help us as we seek to improve farming practice and improve the integrity of our food supply chains. A good starting point might be to apply these as tools for committed producer groups that are already aligned with supply of a premium product or market.
we apply software and models to agricultural data.