Optical Gas Imaging (OGI) has been around for some time, and refers to a class of methods that visualize gases in a scene using highly specialized cameras. It has become popular in recent years due to the improved sensor performance and reduction in sensor cost. In its core are photon detectors - some of which may be cooled to sub-kelvin temperatures - that can capture the slight changes of light as it passes through the gas being imaged.
As a type of gas detection technique, OGI camera systems can provide much more insight to the gas being observed. This is because OGI camera is able to provide a spatial-temporal information of the gas in its field-of-view (FOV). The information picked up by OGI camera is of a higher dimension compared to point-type sensors and laser-type path detectors, which are designed to provide only 1 dimensional temporal density/line-intensity values. In some sense, the data sampled by a laser-type detector represents one single pixel value of an image produced by an OGI camera system. This fundamental difference allows OGI camera to be much more effective in determining the existence of gas in its FOV compared to other point-type and laser-type sensors, and in some situations, also enables source quantification.
As in the case of any sensor device, the performance of OGI camera systems are also affected by the environmental conditions, such as wind, temperature, and humidity. Usually, industrial OGI camera systems are used to detect gases that cannot be easily seen with the naked eye, such as natural gas (with methane as its major content), propane, and many other hydrocarbon gases. If the OGI camera system requires active source of radiation input to 'light up' the scene in order to detect gases, it is classified as an active OGI. Otherwise if it uses environment radiation alone without external radiation source, it is classified as a passive OGI. For any OGI camera system to see gas, no matter whether the system is active or passive, the gas must emit/absorb radiation that is coming towards the camera sensor from its background at a rate that is different from the surrounding air. In the literature, a temperature like quantity called ΔT is often used to describe the difference between gas temperature and its apparent background temperature. In the case where ΔT is very small, the gas is transparent and cannot be detected by the OGI camera system. Thankfully, in most industrial applications, the OGI camera device is positioned to look at a location that is either hotter or cooler than the gas, effectively providing a big enough ΔT for the detection system to work.
Normally, many hydrocarbon gases such as methane and propane are invisible to our naked eyes. That is, they are transparent in the wavelength of light our eyes can sense. However, in other wavelength regions, these gases can absorb or emit light, making them visible to special photon sensors. With special optics and sensors, our OGI camera product is set up in such a way to maximize this effect, and let you visualize the invisible gases.
Since the attenuation of light is related to the properties of the material through which the light is traveling (a relationship also known as Beers' law), we are able to tell how much gas is present in the scene for known type of gases. Cheers all for Beer's law!
Traditional gas sensing solutions can be categorized based on the dimensionality of the data collected. Point sensors (i.e. Catalytic, Point IR, ECC, PID) sample gas concentration at a location , providing 1D data in the unit of density (i.e. ppm). Path sensors (i.e. Open Path IR) detect gas concentration along a sensor path line providing a measure of the total gas present on the sensor path recorded in the unit of density-distance (i.e. ppm x meter). Our optical gas imaging (OGI) solution measures gas concentration in a volume (3D) covered by the camera field-of-view (FOV), and provide an image of N pixels, where each pixel value can reflect the amount of gas present in a volume in using a mass-like unit of density-volume (i.e. ppm x meter3).
In actual outdoor situations, the leaked gas is dissipated quickly into the surrounding environment, one will need to carefully select and deploy a suite of point and path sensors to capture an actual leak. If the leaked gas does not move or dissipate over the sensors' locations or paths, the sensors will not detect the gas leak. Because we rely on imaging optics for gas detection, we are able to detect gas at a distance in outdoor conditions much more effectively.
It is unfortunate that using advanced optical imaging sensors alone is not enough to effectively detect gas leaks. The gas signal is mixed with other signals such as noise from the environment, and one relies on complex mathematical transformations to effectively retrieve the actual gas signal. Accomplishing effective gas leak detection and quantification calls for the power of computation.
Our solution combines edge and cloud computation to bring the best of two worlds to your advantage. We deploy sophisticated real-time computer vision algorithms on the edge for gas plume localization, and feed the edge data back to the cloud for further statistical analysis. This allows us to effectively monitor gas leaks in real-time and provide fast detection alerts to any user anywhere.
With up to 21 TOPS computational power on the edge and superb low power consumption, we can effectively run powerful data-driven models to enhance our autonomous gas detection and reduce the number of false alerts. Our efforts to balance computational complexity and computational resources on the edge and in the cloud have led us to provide a real-time gas detection and quantification solution with reduced operational cost to the user and empower them with the data needed to make the right decisions on the fly.
Named after Leonhard Euler (pronounced “Oiler”), a Swiss mathematician, physicist, astronomer and engineer who made important and influential discoveries in many branches of mathematics. Euler was one of the greatest mathematicians in history, and Euler’s identity: e^(iπ) + 1 = 0 is said to be the most beautiful formula in the world.