Ewald Frch - Fotolia

A digital twin for farming

We speak to a company that has pioneered digital plant growth for indoor farming

At the start of 2020, Capgemini predicted that thanks to advances in technology, vertical farms will soon become a reality. It said that because vertical farms use no soil, little water (compared with traditional agricultural practices), and a combination of natural and artificial lighting, they can produce crops more sustainably than traditional farming.

At the time, Capgemini digital innovations expert Luc Baardman noted: “With companies such as Signify and Valoya continuing to develop better lighting solutions, the costs of running a vertical farm will also decrease.”

Intelligent Growth Solutions (IGS) is a Scottish startup hoping to deliver technology that has immediate benefit to farmers. “Technology has forgotten about farming,” said IGS CTO Dave Scott. “Farmers are so grateful when you give them something simple.”

Scott has an engineering background and said he used to be sent all over the world to fix heavy machinery. Describing how IGS was formed, he said: “Eight years ago, I was introduced to an eccentric farmer who had an incredible vision and an outrageous idea.”

IGS was seeded from this hands-on farming experience and knowledge of farming economics with the tech knowhow from Scott. “If he knew how ridiculous it was, he would never have asked,” said Scott. The idea was to build absolute, real-time control and analysis at no cost.

But when the pair broke down the problem into building blocks, “we achieved 95% of the vision quite quickly”, he said.

IGS is focused on indoor controlled farming, but knowing why certain crops thrive is not well understood. “Light, plant nutrition, air, build-up of gases, water recycling,weather conditions and rain – the combinations are endless and every stage of growth is different,” said Scott. “Plant science is massively misunderstood. We understand more about how a fish works than a plant.”

According to Scott, in farming, a lot of knowledge is in different people’s heads. “Farmers have lots of trade secrets,” he said. “They know nuggets of information about the crops that they grow.”

Working with plant scientists, IGS has developed specific recipes that include variations of parameters such as light, water, nutrients, humidity and temperature. Tweaking these can boost yield, improve quality or deliver a longer shelf-life for vegetables, said Scott, adding: “We are trying things on the fly.”

Rather like a digital twin simulation of industrial machinery, IGS is able to  tweak one or two parameters in the recipe and see the effect on the plant, he said. “The combinations are endless.”

Much of the machinery used by IGS is powered by programmable logic controllers that are designed to run for many years error-free. The company also fits less robust sensors for specific functions that may only be needed for a matter of weeks. “We can micromanage these sensors in our facility in Dundee to improve the recipes in our machines,” said Scott.

Read more about digital farming

  • Retailers are waking up to the potential of vertical farming, with technology-enabled agriculture a core part of supply chain bosses’ and food buyers’ plans.
  • A tech startup industry aimed at modernising farming is making hay in Norway and laying down roots in other countries.

Effectively, IGS sells a combination of the machine for growing plants and a data-driven recipe that uses software to define specific parameters affecting plant growth, to achieve the right outcome for its customers. “Every six weeks, you receive improved recipes,” said Scott. “We provide single-point improvements to the crops you grow via the cloud. Whether you have a machine in Antarctica or Saudi Arabia, you get the same results in terms of crop. You can grow anything with a program that works, but the cost of doing so is different.”

From a technology infrastructure perspective, IGS scans the crops using cameras that capture two-dimensional and three-dimensional images. It also uses sensors to measure water and nitrates. The data is collated in the Microsoft Azure cloud.

At the moment, people are used to analyse the impact of changes to the various plant growth parameters, and for Scott, the next stage is to use artificial intelligence to find new recipes. “We want the machine to look at historical data and suggest the changes we need to make to have a desirable outcome,” he said.

For example, he said, a time lapse of the images taken over 18 months can be analysed with Nvidia accelerated image recognition to understand plant growth.

Clearly there is a cost, which, at least in the short term, is likely to limit how much the digital farming techniques used by IGS, and companies offering vertical farming, can scale. “We’re not going to feed the world,” said Scott, but he sees opportunities in specific niches: “If we can grow proper superfoods with the right minerals, people would see value.”

For certain crops, food miles may also be a factor, where something can be grown locally rather than having to be imported from thousands of miles away, he added.

Other opportunities exist in growing ingredients for perfumes and for certain vaccines.

From the conversation with Scott, another possible application area could be in helping traditional farmers. “Farmers use red diesel and you use a huge amount of it,” he said.

This means that a lot of energy goes into traditional farming. And on top of the energy requirements, said Scott: “The weather is becoming more extreme. Every year there is the risk of complete whiteout versus good enough growth for the supermarkets, which have hard commercial terms. Getting a balance can be very hard to justify.”

Although it may not be on IGS’s immediate roadmap, there are clearly opportunities to use totally controlled indoor farming to run simulations that provide data to help farmers achieve better results from their crops. It is also possible to use indoor farming to develop more resilient seedlings for farmers, said Scott.

Read more on Artificial intelligence, automation and robotics

Data Center
Data Management