A US Oil Pump manufacturer with numerous oil pumps installed in several oil fields wanted to find an intelligent solution to monitor oil leakage/spillage in oil pumps. Their current process involved manual activities of inspecting the images of Oil Pumps captured in the Oil Fields through camera installed at strategic zones in the oil field sending pictures at a 15 minutes interval spanning the set of Oil Pumps.
Due to significant manual activities involved in the entire process, the customer was incurring huge operating costs in monitoring the massive corpus of images in detecting and identifying a possible leak/spillage of the Oil Pumps.
ITC Infotech designed a computer vision solution using Deep Learning technique involving Image Processing and Convolution Neural Networks, which provided the probability of a pump having a leak and also the leak localization zone for an Oil Pump. The solution provided a manifold improvement of the existing manual process in saving both operational cost and time.