Key aspects of digital phenotyping, challenges and opportunities
Aside from being a CHAP member, Gardin is a start-up group based in Oxfordshire focused on developing optical remote sensing technology and machine learning to measure plant physiology and optimise food production.
The renewed focus on sustainability and environmental challenges spurs many efforts towards using digital strategies for crop production. Nowadays people are more demanding when it comes to product origin, mileage and input usage. As a result, there is a real drive from consumers to adopt solutions that protect the planet while providing high quality produce. This trend will help boost digital phenotyping whilst ensuring farmers keep owning their piece of the puzzle and reap the benefits of implementing greener strategies. A win for the grower, the customer and tech companies like us.
Technology adoption is a big challenge – with high costs being one of the main hurdles to overcome. Being able to provide affordable, yet high quality insight is what growers and farmers want. However, it is also important to ensure that accessibility and ease of use are incorporated when developing phenotyping technologies. Finding a balance between these competing interests is challenging, and something we constantly strive for at Gardin.
As technology for digital phenotyping is advancing quickly, it sometimes fails to deliver in a real-world environment. A key aspect is that technology development should be driven by experience – the growers and farmers. Ideally, the user experience, UX, should drive the engineering, and not the other way around. At Gardin we design this principle in our work; and therefore we avoid developing a specific technology riddled with undesirable features, or, worse, lacking the ability to fulfil a crucial function from the grower’s point of view.
Another frustration voiced by many growers is that technology companies place too much focus on overly intricated analytics, which makes them feel bombarded with data. Growers do not want the data per se, they want the tools to solve their challenges effectively. This is where our UX comes in, it aims at understanding the root cause of a problem and supporting growers in developing effective strategies in the farm. Digital phenotyping is multi-layered and complex, but it is one tool that allows you to touch on so many themes.
The use of controlled environment agriculture, CEA, has been of great help to advance and test new digital phenotyping technology. CEA allows us to understand in an efficient manner what happens when you modify a certain parameter or even combinations of parameters all at once.
We observed that the Gardin sensors are well suited for early detection of stress in vertical farming. We basically exploit digital phenotyping to act as a navigation system for commercial growers, alerting them of issues and telling them where to direct their attention. This not only results in a less stressful day-to-day, but it also reduces pre-gate farm waste and quality losses. Even small changes have a massive impact, because they are deployed at such great scale; ultimately we are helping the whole system reach net zero are we are very proud of it.