Illumitex has been known for several years as a supplier of LED lighting products for the greenhouse industry, among others. However, the company’s launch last year of its new FarmVisionAI computer-vision artificial intelligence platform, seemed to set the path for a new journey for the company.
Last week, that transition become official, as Illumitex CEO Jeff Bisberg announced the company is pivoting to become a pure agriculture technology company that will deploy edge vision systems, edge AI, and cloud services to provide visualization and augmented intelligence for controlled environment agriculture.
“With installations ramping, we are seeing that FarmVisionAI is more valuable to growers than LED lighting, and we are refocusing our resources to accelerate deployments” Bisberg says. “By capturing images of every plant at every moment, we create a unique data-set that allows us to close the digital feedback loop and drive amazing new outcomes, at scale, for our customers.”
Greenhouse Grower reached out to Bisberg to get additional thoughts on the move, and what it means for Illumitex and for the horticulture industry.
Greenhouse Grower (GG): What was your reasoning for getting out of the LEDs market? Was it just about focusing the company’s attention exclusively on FarmVisionAI, or were there market trends that dictated the decision?
Jeff Bisberg: First and foremost, the shift is about FarmVisionAI focus and providing digital solutions to help farmers make better decisions faster. While our LED business was growing rapidly, almost 50% year over year for multiple years, the explosion of LED solutions for horticulture into the marketplace creates an intensity and noise to the farmer-lighting-buyer that is an additional barrier to effectively communicating our value message. We were ramping up digital as our main differentiation, and this move really allows us to focus all of our resources on improving the FarmVisionAI digital experience.
GG: What about your existing inventory/portfolio of LED products and R&D? Are these being transitioned elsewhere?
Bisberg: We have fulfilled final orders and have minimal amount of residual inventory to liquidate. We have had some discussions on transitioning our products to other companies, as we have state-of-the-art fixtures that have excellent proven performance in many growing operations, but this is not a key part of the strategy. We are looking for partners on the lighting side that will work closely with us, as it still makes a lot of sense to deploy FarmVisionAI along with lighting.
GG: What is the early feedback you’ve received from growers (specifically greenhouse growers of cannabis, leafy greens, and ornamentals) regarding FarmVisionAI?
Bisberg: Feedback ranges, but because we present the imagery like Google Earth inside the greenhouse, it feels very familiar and intuitive to our users. The reaction that I love the best is, “why hasn’t this been done before?” Visual analysis has always been a central part of farming and now we are just moving it from analog (your eyes) to digital (our cameras). The reality is that folks go blind to all the charts, numbers, and dials of many other digital systems, so we’ll win because we tap into growers’ native behaviors.
GG: AI is a concept we’ve heard a lot about, but to this point, growers may not understand the potential benefits it can provide. What is your plan to address this, and how do you see the potential of AI continuing to evolve in this industry?
Bisberg: Growing plants at scale is very complicated, and it will be a long time before AI replaces anyone. The farmer should think of AI as a new tool that they can use, maybe like a new kind of shovel, to help make their job a little easier, it’s just software, not hardware. Someday it might replace someone, but didn’t the tractor replace people too?
GG: Which industries (in terms of greenhouse production) stand to benefit the most from AI technology?
Bisberg: High-value crops make the most sense, and cannabis just happens to be the highest value. A small positive change in output has a huge benefit that can pay down the costs of new technology. As the technology continues to get better, it will cross over to lower-value crops. Cannabis is the killer app for ag tech, just like that early ear piece mic/speaker was the killer app for Bluetooth. In some sense, outdoor precision agriculture’s failure is that it started with low value crops like corn. It’s just too expensive to improve an acre of corn in the middle of a massive farm to make a monetary difference.
A Closer Look at FarmVisionAI
Bisberg says FarmVisionAI provides three groups of features that can be of high value to an operation:
- Visualization: Growers can see every plant on the farm from anywhere at any time. This provides confidence and assurance as to what is happening to the crop and enables remote high-resolution scouting. Expert consultancy is also available to enhance in the detection and solution of anomalies.
- Augmented intelligence: Horticulturist-curated AI provides accurate digital scouting of what’s happening with the crops. A pipeline of new detections provides continued value and intelligence expansion.
- Labor management: Growers can monitor and track exactly where every person is on the farm by utilizing heat maps, pictures, locations, and activities.
Some of the benefits of FarmVisionAI include:
- Reduced scouting time. Because the system looks at every plant every hour, the user can see at high resolution what is happening on the farm. This can be used to direct more intensive scouting efforts to the areas where plants are in stress.
- Reduced labor. Users can see where their labor has been, for how long, and where they have not been. They can also create achievable benchmarks for procedures from real data, and use that to drive to better efficiencies, while catching areas that have been missed.
- Loss prevention. By detecting anomalies faster, losses are reduced, giving a higher average yield. The time travel feature allows users to rewind the clock virtually and see exactly where, what, who, and how a problem occurred.
- Faster turns. FarmVisionAI sees wasted time and space and sends alerts to create urgency to maximize production in a facility.
Illumitex’s FarmVisionAI can be deployed in any grow architecture from containers, to vertical, to warehouse, to greenhouse. It is easy to install, without the need for motors or tracks, and does not require the deployment of registration markers in the grow. Cameras are wireless and connect to the cloud through Illumitex-deployed gateways and on-site server.
Currently, the system works with leafy greens and determinant plants. While the first AI algorithms are cannabis focused, new detections can be activated with as little as 10,000 images of the desired detection target.