Achieved a 10% production capacity increase for a small medical devices manufacturer by combining machine learning and on-the-line experimentation
Pharmaceuticals
Production Capacity Increase
Client Context And Challenges
Our client, a leading pharmaceutical manufacturer, operates multiple assembly lines producing small medical devices.
The client faced significant production downtime caused by small parts jamming across robotic assembly stations (~50 per line), severely impacting Overall Equipment Effectiveness (OEE) and production capacity.
Increasing market demand required a scalable solution to address downtime and improve productivity.
Our Approach
Conducted a root-cause analysis using manufacturing system data, shift records, and alarm logs to single out machine settings as the main driver of poor line performance.
Due to the lack of historical data on machine settings, we implemented a structured Design of Experiments (DOE) campaign to collect the necessary data for machine learning models to optimize machine parameters on each assembly station.
Packaged the DOE framework into a software interface, enabling operators to implement optimal settings.
Successfully piloted the solution on one line and trained the client’s team to roll it out across additional lines.
Impact We Achieved
Delivered a 10% increase in OEE per line, significantly enhancing production capacity.
Reduced downtime by 50-80% in critical assembly stations through optimized machine settings.
Established a replicable and sustainable process for continuous improvement, empowering the client’s teams to maintain gains and scale improvements across other lines.