“Everyone is measuring the soil, air, humidity, and temperature except for the product itself; the actual plant. Having worked in tech for a very long time, this made no sense to me at all, as decisions were made without having actual product-based metrics,” says Sumanta Talukdar, Founder and CEO of Gardin, a UK agritech company that’s focusing on plant health via their proprietary sensing platform.
“I started Gardin to help CEA growers measure what they’re actually growing at scale. Our solution exists of a patented remote Chlorophyll Fluorescence (ChF) sensor which is linked to the cloud, collects real data on the crop at a huge scale, and shares the insights with the grower to really know what’s going on within the plants,” Sumanta adds.
Knowing the heartbeat of the plant
The Gardin platform has been designed to operate at scale at commercial greenhouses. The sensor is set up in under 10 seconds, after which it operates autonomously. Measurements can be taken by the sensor from up to 3m away, it can measure the entire canopy, including through its height, and take 100s of measurements per hour.
The sensor can also be mounted on a mobile platform, and thus entire greenhouses can be scanned with a single sensor. The raw data is converted by onboard signal processing and then machine learning models in the cloud into easily actionable insights and data for the growers and breeders. The insights can be provided via a front end or via API.
Detecting stressed plants weeks before growers or cameras observe it is a huge win, as Sumanta explains, when it comes to optimizing results and preventing crop losses. Gardin is actively helping its clients with optimizing environmental controls alongside shading strategies, Co2 assimilation, etc., by showing them real-time plant-level responses.
This helps growers iterate and test their strategies, e.g., at a greenhouse level within hours rather than weeks or months. Similarly, the team is helping lighting providers optimize their lighting in real-time resulting in massively lower energy use for the growers and, in other use cases – shorter growth cycles.
Sumanta says, “If you want to understand our product, you just need to look at the team.” The team includes experts in plant science, physics, computer vision, machine learning, embedded systems, data engineering, robotics, electronics, UX, supply chain, and operations. The result is a patented high throughput remote phenotyping platform that Gardin is shipping globally today.
How does it work?
Globally available to any CEA grower, from orchids to vertically grown lettuce to greenhouse-grown tomatoes, the purchase model is rather simple, as Sumanta points out, as it’s subscription-based.
So how many sensors would a grower need per acre? “We always ask the grower what level of control they have on the farm, the rationale being that providing insights at a level that is not actionable is not useful. So if the grower says the smallest unit they can make changes over is an acre, we might say that they need 2 or 3 sensors per acre.
If the sensor can be mounted to something mobile, they will need even less. Ultimately we are an analytics business, not a hardware business, and we are motivated to only sell the bare minimum hardware that’s needed to help the grower.