voc gas leak detection
Real-time VOC Leak Detection For Your Entire Site
Real-time Visual VOC Gas Leak Detection
Without The False Negatives & False Positives
Mobile OGI cameras set the standard for periodic LDAR inspections. Now, get the same OGI accuracy with visual verifications continuously without call-outs
Frequently asked questions
We use an OGI (optical gas instrumentation) camera, like the FLIR G300a, and mount it on a pan-tilt unit. This allows us to cover an entire site with one camera on a 360-degree rotation. The OGI camera is controlled by our AI machine model and the raw camera feed is analyzed every 30-seconds to detect leaks.
We use an OGI (optical gas instrumentation) camera that is designed to be fixed-mounted. This OGI camera has the same core that is used by LDAR teams throughout the world and meets all state, federal and international LDAR standards. The camera is mounted on a pan-tilt unit and is inside of a IP67-rated (intrinsically-safe for outdoor use) housing.
No. Like all CleanConnect.ai machine models, its designed to work autonomously. Our motto is “let the AI do the work,” that way no one needs to be dedicated to stare at security monitors–ever!
If we sold this separately, we would charge $3,500/month for this service. However, you can buy ALL of our AI machine models inside of our Autonomous365 AI suite.
In many cases, clients tried “sniffer” pilot programs that resulted in too many false positives and false negatives. We have seen both in side-by-side comparison tests as well. The sniffers missed leaks because the wind didn’t blow in the right direction (or at all), which led to clients being fined by regulators. The sniffer alerts always have to be confirmed with LDAR teams using OGI cameras. This is NOT the case with our system since we use the same reliable OGI system that your LDAR team uses today to visually detect VOC gas leaks.
A single regulatory fine can cost upwards of $49,000/day in Colorado. In addition,
Yes. We are constantly training our AI machine models to detect even the smallest leaks. One of our clients gave us access to ground-truth-data so that we could accurately quantify leaks to within 1ppm.
Yes. Our AI machine model identifies all leak events and provides a unique ID. We have trained the machine model to recognize a normal event vs. an actual leak. We can examine leaks over time (how long did it leak; what equipment was leaking) so that we can accurately tell the difference and ONLY alert you when there’s an actual leak. Our mission is to eliminate all false positives and false negatives.