DDMRP: Reconciliation of MRP II and the Rise of Demand Driven Planning

Guest authored by Laurent Penard, Chairman and CEO, Citwell Inc.

Emerging nearly 8 years ago, the DDMRP (demand driven material requirement planning) concept and its variations are shaking up the world of supply chain, becoming a global wave that is disrupting established patterns.

However, tightly clinging to their processes, which are deeply integrated into their information systems built around MRP II, many concerned decision-makers are hesitant to question their IT architectures, even though exceptions and flaws are increasingly occurring.

It’s understandable; they grapple with the limitations of first-generation artificial intelligence to the extent that systems are left to operate without the critical input of human intelligence. This unchecked automation renders humans powerless. This is especially true for forecasting calculations that use disaggregation coefficients or statistical algorithms that are not adequately controlled, resulting in unreliable forecasts.

Despite this, the initial MRP II proposal wasn’t just a mirage of technology. MRP II also called for control, simulations, and optimizations with competent human operators. But the machine got carried away, and the mirage turned into a thoughtless machine.

Remember the S&OP brick? Forecasting sufficient requirements to manage purchasing, production investment or resource sizing strategies was no mirage. Aggregate data had to be defined to describe what product families and sales volumes would be in a few months and years’ time, enabling requirements to be calculated using ranges and bills of materials that are sufficiently “macro” to remain modulable, comprehensible, and stimulable.

Unfortunately, the MRP II concept passed too quickly from industrial and flow management practitioners to technology specialists: ERP and APS publishers who turned it into a global system integrating all data and all levels of plant and distribution network management. They brought with them a new promise: Big Data. Millions of SKUs, a myriad of nomenclature, forecasts mixed with hundreds of thousands of order lines - nothing could scare them.

Except that the short-term result of this mad machine was to propose hundreds of manufacturing, purchasing and transfer orders to be launched or readjusted daily. These orders were questioned by the humans behind their screens, as they tried to bypass the calculations, or even to reproduce in Excel, the system calculations that had cost their company tens of millions of dollars.

In the short term, the proponents of automation failed to understand that the whole logic was collapsing, not only on the demand side, which was becoming ever more variable and spread over more and more products, but also on the production side, with resource shortages, supplier disruptions and, finally, a needs-resources balance that was becoming ever more complex to model. Therefore, it was necessary to return to priority management based on real demand, as proposed by Lean with Kanban. Even in the medium to long term, many tools are not equipped with the necessary facilities essential for reflection and simulation around scenarios developing end-to-end logics of high, medium, or low assumptions on demand, industrial and logistical capacity.

A concrete illustration of the weaknesses of demand analysis is the lack of segmentation between demand channels and customer behavior. While demand manifests in various forms, ranging from immediate needs to significant calls for tenders, promotions, regular sales, launches, or events targeting specific customer segments, systems often aggregate and indiscriminately blend these diverse demands together.

THE REVOLT OF THE FLOW SUPPORTERS

The proponents of flow revolted against these first waves of automation, waving the Toyota model in the wind, backed by its results. For example, by preaching the duplication of resources to ensure Takt time in all demand scenarios, and the creation of inventory control labels and needs accumulation tables at each manufacturing station: Kanban and Heijunka.

The revolt could be violent: a factory based on the Toyota model had to ignore orders from the ERP or sales forecasts to manage its flows. Hence the need for considerable investment in resources, and the difficulty of operating in extended mode with distant subcontractors, flows managed to order, or complex internal/external flows. Thus, the Toyota model stood like a wall, irreconcilable with MRP II, even though it would itself fail in many situations (very few implementations crossed factory boundaries).

TIME FOR RECONCILIATION

Even worse, the wall-builders currently in vogue were also out to separate markets, consumers, and industrial sources. As a result, central distribution and planning teams adopted APS methods and tools, which were highly effective if customer and consumer demand remained sufficiently homogeneous and stable.

Naturally, thousands of consultants have flocked to both camps, to the point where a manufacturer or distributor can only transform itself by seeking assistance from the proponents of Lean on the one hand, or from the proponents of forecasting and APS on the other.

So far, neither side has sought consensus, nor been willing to understand or integrate the other’s arguments.

Citwell, a pioneer and leader in DDMRP initiatives, asserts that reconciliation must occur through the recognition that DDMRP serves as a bridge, connecting previously irreconcilable factions.

There are many reasons for this, if you are fully aware of all the aspects of Demand Driven concepts, without making reductive judgements such as “your demand driven is nothing but red-yellow-green buffers”. Although we must never stop explaining, our response to judgments of this kind is to send their authors to certified training courses, which are very inclusive of all concepts and business issues.

So, in 6 points that are no doubt a little reductive, here are some reasons why reconciliation will take place, leading to a global theory, intelligent systems and humans in control:

• Kanban brought visual management based on real demand > DDMRP digitalizes it and makes it visible to everyone at the same time
• MRP II creates decoupling points in BOMs > DDMRP structures and simulates them
• MRP II produces automatic reports on orders to be placed > DDMRP makes it visual and stabilizes its operation
• Heijunka facilitates scheduling > DDMRP integrates bottleneck scheduling logic and standardizes product/ campaign wheels
• MRP II helps build S&OP > DDAE (Demand Driven Adaptive Enterprise) makes S&OP more relevant by segmenting demand and focusing simulations on constrained resources
• LEAN promotes flow without equating the notion of full cost > DDAE focuses on a vision of overall cost and margin achieved on bottleneck resources.

Yet, achieving reconciliation will necessitate concessions from both sides, acknowledging systemic and organizational choices rooted in a bygone era.
 

"Citwell, a pioneer and leader in DDMRP initiatives, asserts that reconciliation must occur through the recognition that DDMRP serves as a bridge, connecting previously irreconcilable factions."

Neither MRP II nor Kanban proposed methods for monitoring and optimizing parameters, on the assumption that forecasts would be increasingly reliable and mathematical formulas sufficient to size stock or optimize scheduling. While Kanban quickly reaches its limits when it comes to managing multi-site exchanges involving thousands of products, APS and other integrated solutions should not overshadow the need for collaboration between the various links in the supply chain.

Organizational silos will be broken down not only by integrated systems, but above all by the trust that will be built up through visibility of service contracts between suppliers and factories, as well as between factories, distribution teams, and sales subsidiaries. Citwell and Demand Driven are therefore committed to reconciling the best of each approach and are engaged in profound transformations that succeed by challenging obsolete dogmas and overturning the paradigms that have led to the failure of the service cost-stock-management strategies.

It also involves the judicious application of the best of change management, data analysis, artificial intelligence, and visual performance management.

Our ability to implement these practices, without being dogmatic, by listening to the culture and the needs of our customers, is the key to our success of our customers’ skills is at the heart of the DNA of our transformation processes and of our customers’ success.

At the same time, as supply chains become more complex, traditional MRP II systems are increasingly falling short. DDMRP provides a more adaptable, demand-driven solution, improving inventory management and lead times through a focus on real demand and strategic buffers. To visualize and learn more about how DDMRP can transform your supply chain, consider joining our upcoming DDBRIX workshop for practical insights and guidance.

https://www.citwell.com/us/events/citwell-at-the-ddbrix-tour/