Miller Magazine Issue: 127 July 2020

57 COVER STORY MILLER / JULY 2020 sources are determined and eliminated. 3- Predictive maintenance (PdM) measures manually or by online sensors parameters of machines and equip- ment during working. For example, measuring of eleva- tors bears or mill rollers bears temperature. The data are recorded by specialized software (Computerized mainte- nance management system – CMMS) to plan a mainte- nance schedule. This software traces data and historical data using statistic analytical approach to highlight where a machine is not performing as it should so that it can be repaired ahead of time. Predictive maintenance can help mills make repairs when or before they’re needed, rather than running breakdown maintenance or replacing a per- fectly good part that may have many cycles left (Preven- tive Maintenance), thus reducing costly downtime. PREDICTIVE MAINTENANCE CHALLENGES Effective predictive maintenance practicing is faced by following: (1) High demands on data access. This challenge is faced by new technologies sensors that made it is now possible to read parameters of equipment parts, like tem- perature, vibration, relative position, tightness of driving belt... etc. (2) The capability to deal with industrial big data. Big Data describes the huge amount of collected data from sensors. Just big data has ability to analyze and can be used to find trends and to determine cause and effect, and its implications for decision-making. (3) The prediction accuracy of predictive maintenance. After collecting large amounts of data, it is very import- ant to use the data to take a decision about predictive maintenance. The inaccurate decision may result in either unnecessary maintenance, or production downtime be- cause of unexpected machine failures. HOW MAINTENANCE IS TAPPING INTO INDUSTRY 4.0 Industry 4.0 was presented in 2013 as the vision for the next industrial revolution. This was due to great advanc- es in recent technologies and their potential to improve manufacturing environments. Industry 4.0 described the transition way to Artificial intelligent “Smart” industry. To achieve industry 4.0 in the mill, three principal as- pects “Magic Triangle” have to implemented in the pro- cess: 1. A huge digital database will be created from the re- sults that have been read by advanced high technology sensors. This data collection has to be stored in a cloud