Flow Profiles vs Inventory Buffer Profiles
For DDMRP inventory management, buffer profiles are defined. A buffer profile groups together articles with similar characteristics – replenishment times, demand variability, supply variability. This will allow to assign identical buffer sizing rules to all these articles (red, yellow, green zones, detection of demand spikes, etc.).
When it comes to building a Demand Driven Operating Model (DDOM), we will identify, in a similar way, flow profiles. For a set of items that follow similar flows, we will apply common scheduling and execution rules. These steering rules define:
- Where are the critical control points in the flow – the control points
- How to plan, at finite capacity or not, and how to promise a reliable availability date
- Where to place time buffers, and how to size them
- Which grouping and sequencing rules to apply
A flow profile will therefore group together a set of products with similar routings.
Depending on your type of production, this exercise will be more or less simple. In any case, you will have to mobilize a team that knows your environment well: production, methods, planning, quality…
Macro Value Steam Mapping
The approach is similar to establishing a standard VSM, but generally at a less detailed level. You are not looking for waste or the ratio of non-value added to value added. You are looking to understand the flow mapping to implement a pull flow mode that will then allow your teams to improve.
What you will be looking to identify in this analysis is:
- What are your constraints. The steps that define the pace of your flows or branches of flows.
- Where are queues happening.
- What are the main routes that your products follow.
- What variability is observed: for example, the average, median, minimum, maximum time and standard deviation between two process steps.
You can use team knowledge, a gemba walk, but also data analysis, which will allow you to factualize and accelerate the analysis.
Demand Driven AND Data Driven
A very useful approach in this context is to use process mining techniques.
It is a matter of exploiting the transaction histories that you may have in your ERP or your MES system, to draw a flow map, and quantify the frequency of use of each resource, its workload and the waiting times between operations.
This technology, which was still confidential and expensive not long ago, is now becoming more accessible. There are multiple solutions on the market, for example cloud solutions or Power BI visuals. A little research on “process mining” should give you some clues.
Visiting the workshop with the project team and designing the steering model based on this mapping will save a lot of time and allow you to reach a consensus very quickly.
Marrying Data Analysis and Collective Intelligence
If, for example, you notice that WIP often accumulates in front of a given work center, that this process step is one of the busiest, and used in a large number of work orders, these are indications that this may be a constraint, and therefore an ideal control point.
If you notice that many orders pass through a resource that is not necessarily constrained, but that constitutes a crossroads of flows (a paint line, the quality lab, etc.), this is probably a control point in front of which a time buffer should be inserted.
These process mining tools will also help you identify flow families by detecting variants and analogies. For example, some flows share the same primary constraint, some include several passes on the same primary constraint, involve more than one constraint, or no constraint at all.
On this basis, it will be easy to transcribe the control model into our DBR+ solution, for a digitalized pull flow adapted to your environment. We will come back to this in future articles.
Want to explore this in your context? Do not hesitate to contact us!
