The Possibilities of Improving Maintenance Through CMMS Data Analysis

Abstract

Maintenance of production equipment is one of the most critical support actions in manufacturing companies for staying competitive. More recently, with the introduction of Industry 4.0, academia, as well as industry, put a lot of effort into condition monitoring in order to implement predictive maintenance. Most stakeholders agree that maintenance need to be more data-driven. However, in order to draw true advantage of data-driven decisions, it is necessary for manufacturing companies to have implemented basic maintenance to a high standard in order to reduce for example: recurring failures, human errors, unsafe machines, etc. The real-time data can then be used to improve efficiency of maintenance tasks and schedule that adds value to the processes. In manufacturing industry, maintenance actions are commonly administered in a Computerized Maintenance Management System, CMMS, still, rather few companies analyze their maintenance records. Behind these data there is often a treasure of improvement opportunities that could be used to improve basic maintenance. The purpose of this paper is to explore how historical data from a CMMS can be used in order to improve maintenance effectiveness and efficiency of activities. In order to exemplify the possibilities of analyzing CMMS records, a case study has been performed in a plant, manufacturing driveline components for heavy construction vehicles. The study shows that one major obstacle for utilizing the CMMS data is poor description of faults and failures when it comes to work order requests, mostly performed by operators and assemblers, as well as work order reporting, mostly performed by repairmen and maintenance technicians. However, by thorough analysis of well described corrective maintenance, it is possible for industry to understand the nature of the occurring breakdowns and thus, refine the preventive maintenance program in order to further increase the dependability of the production system.

Combining Ontology and Large Language Models to Identify Recurring Machine Failures in Free-Text Fields

“…Perceived by employees at the company and to a certain extent analyzed [9], the case company is suffering from recurring failures. To identify these recurring failures, maintenance engineers engage in a manual analysis of work orders within the CMMS.…”

“…Also, it has been reported that reusing experiences as reported by maintenance employees are limited [8]. To improve this situation, companies are advised to analyze their current state and the historical behavior of their equipment [9]. Employing a Computerized Maintenance Management System (CMMS) is essential as a tool in this analytical process [10][11].…”

Data-Driven Approaches for Operation and Maintenance Decision on Power Plants

“…By documenting failure reports, disturbances, planned preventive maintenance, etc. in CMMS, the maintenance efficiency and efficacy of the components have been enhanced (Salonen et al, 2020).…”