CPG manufacturers use predictive maintenance analysis to reduce downtime

About client 

Our client is a leading manufacturer in the CPG Goods industry, specializing in the production of high-quality products like household goods, food, and skincare products. They have set up state-of-the-art manufacturing facilities in many global locations all over the world, with a commitment to deliver products that are consistent with quality, reliability, and safety. 

Challenges 

Their manufacturing had a few major challenges related to equipment maintenance. They are currently using corrective maintenance, where whenever there is a failure, they turn down operations until the technician fixes the issue. 

Production cost wastage: Breakdown and repair of equipment led to increased maintenance costs and costly repairs. They also faced unexpected machinery downtime, where the machine goes down without a warning.  

Workload fluctuations: Client is unable to meet the demand reactive maintenance and not fixing issues on time. 

Low equipment effectiveness: Varying loads and unbalanced maintenance loads across the machinery causing some machines to overwork and some being underperforming. 

Diminishing machine health: Unoptimized loads without timely break is affecting machine health and lifetime, reducing ROI. 

datakulture Solution 

Our data analytics team assessed their day-to-day operations, optimal machinery conditions, and manufacturing goals to deduce a Predictive Maintenance Analysis solution. To visualize the solution, we developed a dashboard around it that demonstrated the following metrics in near-real-time. 

Operational and idle hour analysis: Total hours of operational hours & idle time for each machinery. 

Maintenance schedules: Tracking maintenance schedules of every machinery along with future projections on when the next maintenance sessions will be required. 

OEE (Overall Equipment Effectiveness): Deriving and visualizing equipment efficiency and highlighting them in ‘high’, ‘medium’, and ‘low’ priority levels to ensure that machines in critical condition get prioritized care. 

Analysis on diagnosis: Using three categories ‘normal’, ‘warning’, and ‘critical’ to express machine health based on real-time diagnosis reports. 

Anomaly indicators: Analyzing machine performance trends and highlighting any anomalies noticed, plotting both prediction scores and anomaly values to alert ahead on predicted failures and downtimes. 

Impact the above solution has created 

• Predictive maintenance insights brought unplanned downtime down by 30% 

• Planned maintenance schedules reduced maintenance costs by 25% 

• OEE increased by 10%, ensuring a more stable production schedule. 

• Timely prediction of machine failure and fix reduced faults by 15%, promoting business continuity. 

Conclusion 

Our client’s operations team saw a great improvement in the way they functioned, all thanks to predictive insights and timely intervention. They not only increased their cost savings but also helped them increase equipment efficiency and lifetime, while getting more work done to meet demands. They didn’t have to wait until the machine broke down but rather be more proactive with their maintenance and care schedules, picking trends from data.