Exploring Effective Approaches To Minimize Downtime In Final Assembly Line Of Braking Systems

Abstract

The global automotive manufacturing industry's growth in downtime reductio is substantial, valued at $3272.6 billion USD with a 3.01% growth rate.This growth in downtime reduction underscores the industry's commitment to enhancing efficiency, quality, and overall productivity across its diverse range of operations. Downtime in the braking system assembly line can lead to utilization loss or technical availability loss. In this context, many proactive maintenance strategies are explored but there's limited focus on addressing error-prone machines and utilizing Machine Learning for predicting downtime. The objective is to prioritize downtime reduction through error analysis, critical machine identification, and implementing ML-based solutions. This comprehensive research delved into the significance of minimizing downtime in the braking system final assembly line through meticulous data analysis, visualization techniquesand targeted interventions and then identified key issues and achieved tangible improvements in operational efficiency. The analysis revealed significant findings and then employed the Pareto Principle to identify top downtime machines, illustrating their distribution through a Pareto chart of machine defects. Furthermore, Exceptions and Problem Areas were identified utilizing statistical process controls, offering insights into critical error contributors. Notably, a comprehensive exploration of the most prominent downtime machine was undertaken, evidenced by LCL and UCL charts and a Fishbone Diagram detailing causal relationships. The research leverages real-world data involving dates, machine names, and downtime durations to develop a predictive model that aids in proactively managing production disruptions. The application of RPN calculations before and after error correction demonstrated a substantial reduction from of 432 to 75, validating the efficacy of the corrective actions. The real-time data was used to build a model that can predict when production machines downtime might happen. This helps us be prepared and manage any possible disruptions in production.The outcome of the project highlights the intrinsic link between downtime reduction and assembly line efficiency, emphasizing the importance of data-driven interventions. This culminated in the resolution of key issues, illustrated by the mitigation of the PCBA Pressin machine and LVDT sensor errors, yielding tangible reductions in downtime and notable productivity improvements. In this direction, exploring AI-driven predictive maintenance holds immense potential for advancing downtime reduction strategies. Leveraging AI algorithms to analyse live data streams from machinery and sensors can enable the detection of patterns indicating imminent failures and can pre-emptively prevent downtime.