Case study Production optimization

Optimize to increase equipment availability by 5%

Process optimization facilitates superior automated assembly cycle times, benchmarks operational production, and eliminates variances in product quality

Service:
Other

Improvements in production scheduling

The EMS usually recalculates cycle times for its production planning once a year. This is mostly because the data analysis behind adjusting cycle times has proven to be time and labor intensive. Moreover, such analysis produces only point estimate times that give no insight on variability and interdependencies. Relying on real-time production data from MES, and planning data from the ERP, we produced an analysis that continuously identifies products and operations with the greatest potential for efficiency gains.

Additionally, we dug deeper into analyzing the efficiency of the workforce. Operation times seemed to vary greatly between operators, and there was perceived potential in identifying efficient peers and capturing their best practices. Based on operator work bookings and product complexity, our analysis tool identifies operators who worked fastest with the least variation in quality.

Our tools of wizardry

When we looked at the total time savings, there were a few products and operations that stood out in particular. The most savings occurred when variance in functional and integration testing was decreased.

On SMA line it became apparent, that there is some excess capacity hidden into cycle times. In general, the planned cycle times per component exceed the actual median cycle times.

When looking at worker efficiencies, it became apparent that there exists a small subgroup of efficient workers, who operate at low time bookings and whose work times are very stable (i.e. no high outliers). Those workers should be used as trainers, and their work practices should be recorded.

Recommendations for the client

  • Benchmark operator time for manual assembly phase at approximately 2.5 seconds per component.
  • Look into sources of variance in operator work, booking times in pre-handling and final assembly phases.
  • Implement adjusted cycle times for FCT. The time required to complete this step is sometimes overestimated.
  • Examine variances in ICT cycles. Lack of data makes it impossible to explain high medians and variances
  • SMA cycle times are accurate, and current planning figures tend to overestimate the line’s full capacity.

Methods used

  • Microsoft R Server for .csv files, data.table
  • SpectX for log/blob parsing
  • R with H2O

Contact us

Share your challenge with us

Schedule a discovery meeting with us to share your challenge and benefit from our technological product development expertise to plan your next step.