AL-DRIVEN PREDICTIVE MAINTENANCE IN MANUFACTURING: ENHANCING EFFICIENCY AND REDUCING DOWNTIME

Abstract

Background: Predictive maintenance using artificial intelligence is gaining importance in the manufacturing industry the opportunity to bring improvements in performance and time loss. However, research studies have revealed variation in the extent of improvement across industries and its firm adoption, this calls for a detailed analysis of the impact of improvement on the manufacturing sectors.

Objective: The purpose of this work is to establish the extent of AI-driven predictive maintenance and its effect on productivity, equipment downtime, and maintenance cost within the manufacturing industry, as well as factors that may affect the satisfaction and adoption of the technology among manufacturing personnel.

Methods: The cross-sectional survey research design was adopted with the use of constructed questionnaires to the respondents from manufacturing companies. AI satisfaction and adoption rates were also captured alongside high-level KPIs that include reduced time losses and containment of costs. Frequency distributions and percentages were also calculated on the data obtained Using statistical tools such as correlation analysis, regression analysis, and reliability analysis using Cronbach’s alpha coefficient.

Results: It was further established that the respondents have a positive attitude towards AI-driven predictive maintenance with a high probability of recommending the approach. However, the Shapiro-Wilk test was used to determine normality in the response and the results pointed towards diversity. Cronbach’s Alpha indicated poor internal consistency of variables such as satisfaction and agreement with AI impact hence indicating variability in perceptions. Little correlation coefficients between key variables indicate that antecedents to satisfaction and perceived benefits are complex.

Conclusion: Predictive maintenance through artificial intelligence has the potential to enhance the various manufacturing operations; nevertheless, its usage and reception of effectiveness feature a dissimilar pattern concerning the different workforce categories. Satisfaction was higher among the bank’s full-time workers, whereas satisfaction among part-time workers as well as retirees was more variable. Incorporating a wider population of workers and improving the assessment methods used are some of the factors that would require further research in the future.