Our platform uses a forward facing neural network and input from an omnidirectional ultrasonic microphone array to pinpoint compressed air leakage from your system.
Any fluid that is pressurized that escapes through an orifice will produce ultrasonic sound.
As air escapes through non-productive leaks, it produces ultrasonic sound we can use to locate it. Once the potential leak is found, a thermographic image is obtained to detect the high pressure fluid escaping. As the air escapes, it expands, causing it to cool down.
Using computer vision, we can then pinpoint the leak and tag it in our system, for it to be fixed by a human later.
Currently we are working on integrating our platform with popular ERP systems such as SAP to automatically create a service ticket and make for a speedy repair.