Sramana Mitra: What’s interesting to me and created a bit of cognitive dissonance is when you think of gas, I would assume that you’re checking for smells.
David Conley: The OGI camera has been around for 15 years. It’s a cold core mid-wave camera. It’s filtered to 3.3 microns. It uses a notch filter to see the energy between the backgrounds of the pixels and the camera. You see the pixels and then you’ll see the energy between them. If you filter to 3.3 microns, you’re seeing gas. Every VOC will fall into that with a couple of exceptions. You’re seeing at least 24 VOCs simultaneously in one view. That camera already existed.
The problem with it was the way that you operated it manually. You had to send people in a truck.
The number one safety issue in most industrial spaces is accidents. That’s not even including the emissions reduction by taking trucks off the road. They had 650,000 scheduled LDAR inspections in Colorado alone. A percentage of those resulted in not only safety incidents but also emissions. Out of those 650,000, there were 2,400 reportable leaks that they found. That’s less than 2% of the scheduled callouts that resulted in any progress.
The regulation is counter-intuitive. They’re trying to slow emissions by manually this creating more emissions. We just replaced it with a robotics AI solution that’s using the same technology that they trust and use. The model itself for gas detection is a deep learning model. It’s a segmentation model.
Sramana Mitra: The company that approached you showed you this problem.
David Conley: Yes. This is another teaching moment. Make sure you get contracts. For the first customer that funded us, we set a success criterion. If they met the success criteria, then they would buy it. We did front the money to them to build the initial proof of concept. As soon as we hit the success criteria, they bought it. Then they said, “Now we want 10 other models.” They cut us a check for a couple of million dollars and we’re off to the races.
Sramana Mitra: This client sounds like a big client and you’re doing a lot of work with them. How many clients do you have?
David Conley: We have three different clients from POC clients all the way to Scale Clients. Between all three, we have around 20 projects, all in different stages. We have five scale clients. We’re on the journey right now from Minimum Viable Product (MVP) to Minimum Viable Repeatability (MVR). The next metric we want to hit is MVR. The revenue skews because of the vertical that we’re in. It’s expensive to run our stuff. It’s not super novel. Our revenue is a lot higher. We’re doing $6 million a year. That’s great but we’re still not MVR. We need to have 20 scale clients.
Sramana Mitra: If you’re operating within oil & gas, how many companies are there that can sustain scale deals?
David Conley: The number of customers are not very many. As for the total number of locations, there’s a ton. Each customer will operate a couple of thousand locations. The definition for scale for us a hundred locations or more. You’re right. The number of customers is probably in the thousands. Probably less than 10,000. But the number of locations they have is massive. We started with the highest-end product.
