Sumanta Talukdar is changing the way we grow crops

Although his business model is heavily tech-based, it started with two premises: first, why isn’t there a way of measuring the nutritional value of crops, destined for food products, as they grow? And second, that any disruptive business in the agricultural sector has to work within the physical parameters of the way farming works now, or the industry would not be interested.

Gardin is an agri-tech start-up, founded about three years ago and based in Oxfordshire, South-East England, with a business model that focuses on using AI to examine, analyse and optimise nutrition value and yields of crops. It does this while maximising efficiency by using fewer resources, and measuring the health of the crop rather than simply appearance. Talukdar has proven to clients that “we can help them spend less money while improving their operations”. Their primary focus is on growing crops, from tomatoes to berries, in intensive environments, including vertical farms.

When it comes to measuring crop yield, it was a personal choice and obsession with “measuring” for Talukdar for the creation of his second start-up company. Utilising data sets and analysing the results is crucial for bettering performance generally. Talukdar explains this results in two key streams: “you can, a) make better products, and b) you can optimise how you make that product”. Applying scientific methodology to the food industry is something that Talukdar believes is missing from the field currently and is an element he seeks to rectify with Gardin. 

Analysing the crop data is cited by Talukdar as a critical factor of the business’s success, and showcasing their results in real time. Talukdar asks, “Should you not be measuring the health of that crop, because a healthier crop should naturally lead to a better outcome?” He jokes, likening Gardin to a doctor, “We take the pulse of the plant”.

Diving deeper into the evaluation of crop health, the preference for phenotypic observation over cosmetic observation is explained on Gardin’s platform, consisting of a proprietary sensor and machine learning. Gardin measures growth, tolerance to stress, pests and disease, genetic and plant diversity, and population. This is a large accumulation of data, especially when compared with the industry-standard measure of plant health, the Leaf Area Index (LAI). This mode is effective, but also time consuming, labour intensive and does not give the full picture.

This use of AI is what Gardin believes sets them apart from competitors in the field. “We have technology competitors that we know of, but none operate on the same software-as-a-service business model.” Talukdar expands by explaining that efficiency gains from using their software are proving particularly important at a time of rising energy costs, as many clients use heat and light for their intensive farming.

Clients are predominantly focused on horticulture and protected cropping, from polytunnels to greenhouses and vertical farms. Gardin works closely with them to “find ways of improving their produce, whether that is increasing yields, or uniformity, for example”. Talukdar notes that greenhouses, in particular, are struggling with energy costs, so they are working with many to find a solution. As the company creates viable solutions for their own clients, what happens to the yield of crop producers who are not clients? Does the gap continue to widen within horticulture or does everyone start using Gardin to level the playing field?

Talukdar shares his observations: “if you have two greenhouses owned by two different companies, right next to each other, what we’ve realised is that they don’t view each other as competition. They actually cooperate with each other more, because they have a common challenge with the retailer. They are not natural competitors.”

The varying skill set required for such an innovative enterprise requires a diverse set of individuals. “The team is the engine,” Talukdar outlines. “This is a full-stack product, meaning it involves everything – hardware, software, machine learning… When you look at the team, it looks like the kind of team you need to build a product like this.”

When asked about the future, and whether he has his sights set on the rest of Europe, Talukdar replies, “it sounds like a nice problem to have, but our challenge is going to be getting that balance of growth right, avoiding unnecessary expansion”. The aim is to “actually grow and make sure that we can keep supporting our customers. If they work with us, we can help them effectively make more money, using less resources.”

Biography / Key highlights

2004 Completes a PhD in photonics at the University of Sussex

2014 Co-founds his first start-up, WaveOptics, developing see-through augmented reality display technologies

2019 Founds his agri-tech business, Gardin