The purpose of my thesis is to demonstrate the infrastucture of a system, which is capable of using Predictive Maintenance. The idea behind Predictive Maintenance is that, processing data from multiple sensors, which are attached to a device, can provide relevant information about the device’s condition. The advantage of using Predictive Maintenance is shown in the decreased number of necessary maintenance of the monitored device, which is achieved through well-planned maintenance schedules based on the conditions of the devices. This way the utility of the device can be maximalized and the loss of the company minimalized. The loss originates from the profit deficit caused by stopped conveyor lines.
My thesis consist of three difference parts, which are the data gathering, the data forwarding and the data processing. The description of the first part ranges from reviewing the features of the used sensors and the method of data gathering through the uniqueness of the collected data structure. The second part demonstrates my solution for the Gateway and the process of storing data to a cloud. The third part contains the description of the collected data’s processing. Before the last part I take time to show the idea behind the Predictive Maintenance in further depths and mention few options besides it.