Difficulty in identifying bottlenecks in high throughput assembly lines as the process is non-linear in nature.
The Before State
Cases processed per hour were at a very low value of ~400/hr as against the expected target of 800/hr and the root cause of the problem was not known.
Utilized Process Mining to identify bottlenecks like:
- 18% of cases stuck at carton sorted area
- 20% of cases diverted back to carton sorted area from the aisle lift pickup as they arrived at wrong aisle
The After State
Cases processed increased to 680/hr/door
- Inventory levels reduced from 48 to 45 days
- Savings of $1 million per packaging line