Hello! I’ve been experimenting with Global Scheduling at our site for the past few months and have been struggling to find a way to truly minimize WIP while running finite capacity on our front-end operations (machining).
Our BOMs are very complex and can contain anywhere from 20 to over 100 subassemblies, with total lead times ranging from 4 to 12 weeks.
To reduce WIP, I’ve tried several approaches, including:
- Running Global Scheduling with Minimize WIP both enabled and disabled
- Reducing our planning horizon from 365 days to 90 days
- Using scheduling blocks with split jobs checked
- Enabling Do Not Allow Scheduling in the Past
- Assigning larger jobs a High priority and smaller jobs a Normal priority
- Increasing capacity to make more room in the constrained resource gorups
While scheduling blocks provided a noticeable improvement in capacity utilization, none of these changes have significantly reduced the number of jobs that begin building WIP far in advance of their actual need date. It seems global will plan operations, whether constrained or unconstrained, too early.
Some extreme examples are shown below:
-
Assembly 1 is scheduled to start in September but the parent isn’t due until February
-
Assembly 2 is starting too early as well as 3,4,5 & 6
If I were to rerun these jobs in Job Entry against finite capacity and minimize WIP, they would schedule more reasonable. I’ve also tried using the load leveler but due to our complexity, it just crashes after hours of running.
Thanks in advance for the feedback.