A new expression – digital twinning – is becoming increasingly common when talking about AI and agriculture, it has now taken a broader meaning with the rise of hybrid events that operate in real and online environments.
Digital twinning describes how virtual versions of real-world systems and situations can be created and modelled to enable different scenarios to be tested. It is particularly valuable for ‘what if’ type questions as the system can be pushed to its limit to identify points of weakness, inefficiency and explore its response to such changes.
We are increasingly experiencing our own ‘digital twin’ experience as we try to replicate the intangible benefits gained from networking in an online world.

Online versions of real world events
June saw us deliver a web-based Pollinator on ‘alternative and novel crops‘, host the final of our GROW agri-tech business plan competition, attend online Cereals 2020, and create a virtual Innovation Hub full of agri-tech solutions.
The direction of travel for conferences and meetings is towards a “hybrid” format where smaller scale real-world activities run with a simultaneous online audience. This requires very different planning and delivery to serve the different groups, hence the so-called “hybrid” model.
Harnessing the power of data
The concept of digital twinning has been gaining traction across all industries, with Ocado being one of the first to create a digital version of their warehouses for the food sector. The company also uses digital twins upstream to forecast customer demand and downstream to model the algorithms for van routing. It says it is on a journey to build an end-to-end digital twin of its entire ecommerce, fulfilment, and logistics platform.
The real power of digital twinning is harnessed where it is based on a specific, live situation, and crucially, is informed by real-time data.
Data about specific fields, individual animals, buildings, farms, landscapes and entire supply chains or systems, is combined with other data sources such as weather or commodity pricing, and used to train the AI algorithms and create an environment that replicates the real world.
This approach de-risks the costs and potential hazards of different real-world interventions by being able to run models and different scenarios virtually (and hence cheaper and more safely) rather than in real life. It can also help predict outcomes.
Of course, setting up a digital twin of even the simplest thing is complicated, potentially expensive and is heavily reliant on reliable robust technologies. (As we found in our small, modest way at the Innovation Hub!). It requires a deep understanding of the real-world system to ensure the digital system mirrors the live one.
Collaborative working
But the power of the tool to help end-users and developers work together is not to be under-estimated. Similarly, digital twins can also help inform policy, support decision-making and guide investment decisions, and ultimately even on-farm decisions.
The concept of “hybrid vigour” – where the resulting blend is stronger than both parents, is well known and is one we hope to apply to our future activities.
References to twins in literature, mythology and legends often are linked to farce, confusion, sibling rivalry or even murder. Happily the real-world experiences of hybrids and twins – both real and digital – is likely to be more positive and feature more in our future.