This research study proposes a new alternative for monitoring and managing maintenance work in neonatal respiratory equipment at the Edgardo Rebagliati Martins General Hospital, a general health facility in Lima, Peru. This proposal is based on a Mobile Vision Optical Character Recognition (OCR) algorithm, which identifies the numerical code of each device in the field through an Android mobile application. This information is then processed using a Weibull model to determine equipment failure probabilities. Based on the mean times between failures collected from history information, the system calculates the current failure probabilities for each device. These probabilities are recalculated every time the equipment is scanned. This paper describes the variables used, obtaining an 88% accuracy when using the OCR system, thus identifying preventive maintenance needs from device information.