Predictive maintenance in LED street lighting controlled with telemanagement system to improve current fault detection procedures using software tools.

Abstract

Predicting the lifetime of LED light sources becomes quite challenging because the time to failure is long. The LM-80 and TM-21 methods are the main used by companies to establish the product lifetime. Accurate the RUL prediction can facilitate predictive maintenance. Predictive maintenance allows estimating when a failure will occur. In this context, the maintenance can be planned in advance, eliminating unplanned outage and maximizing the useful life of the equipment. In this work, the LM-80 and TM-21 methods are used for the acquisition and extrapolation of luminous flux data, which are entered into an algorithm developed from an exponential degradation model. With the result obtained, it is possible to establish actions that allow predictive maintenance in LED street lighting controlled by a remote management system and achieve a longer service life.