Job shop operations involve the conversion of inputs to outputs in environments characterized by low volume orders, custom builds, scarce machines, scarce people and scarce materials.
Balancing supply with demand is key for any business – job shop operations are no different but there are a number of unique challenges in job shop operations not present in high volume manufacturing operations.
Given a specific infrastructure (e.g. premises, equipment, people, access to materials, and access to working capital) in the presence of a steady stream of orders, the key to higher profits is to make efficient use of this infrastructure.
At the end of the day it’s all about throughput and quality control.
If machines and people are not kept busy you have under capacity whereas overloading stresses the organization and can cause quality to suffer (e.g. machines break down, people make mistakes), so it seems logical that keeping things running at a reasonable pace with a focus on minimizing time gaps between steps is likely to produce the desired result (i.e. good use of infrastructure).
Putting these ideas into practice for job shop operations requires two initiatives:
- When orders come in, processes need to be immediately mapped. It has to be easy and quick. (i.e. graphically map out the steps, identify the resource needs at each step along with any forms needed to collect data and to provide evidence of completion of work).
- Roll out each mapped process as a template for the order to a 3-tier RALB (resource allocation, leveling and balancing) scheduling environment and let the environment do most of the heavy lifting;
Here’s how throughput can be increased.
Level 1 Scheduling
The environment auto-posts steps to the attention of the right people using logic connections in your maps, plus resource tags at steps.
Workers perform steps, fill in any forms at these steps, and commit steps.
Soon as a step is committed process map logic kicks in again posting the next-in-line step to the attention of the appropriate workers (no unwanted time delays between steps).
Level 2 Scheduling
Let’s do a reality check here – It’s fine to eliminate time gaps between the time tasks are committed and the posting of next-in-line-steps.
However, workers rarely focus on one order at a time. (i.e. machines may be busy, materials may be on backorder, a needed assembly may not be ready for consolidation with another one) so we need to let workers micro-schedule their own work, going off one order to another and then later returning to a suspended order.
Level 3 Scheduling
Next we have imbalances that are the result of any number of things (changing customer priorities, failure of an assembly at QA, receipt of a change order) any of which can result in a need for supervisors to level and balance workload across machines and people.
It’s hard to argue against the logic of the 3-tier scheduling approach to workflow/workload management for job shop operations.
What’s in your scheduling algorithm?