How a parameter-driven approach to value chain planning helps to solve the planning riddle
While the fully autonomous, AI-driven value chain remains a promising vision, we believe that a parameter-driven value chain planning approach is a viable and important step in this direction. Besides, a parameter-driven value chain ticks many boxes to tackle today’s planning challenges leading to more robust and resilient supply chains – this is vital to coping with global challenges like COVID-19. In a series of four articles, we will explore how parameter-driven value chains work, what benefits they drive, and what this means to the management of your company’s value chain.
In this first article, we will explain why a parameter-driven value chain is relevant and important.
Managing value chains is getting ever more challenging
As CAMELOT partners Thomas Ebel and Christian Kroschl describe in their blog article “Calling for a new value chain model” the increasingly volatile and uncertain external environment forces businesses to fundamentally re-think their value chain models. New planning approaches are one of the central pillars of future value chain models. Also, the shift to new planning approaches is an essential cornerstone of achieving greater value chain resilience. The article on “Supply chain risk management – the rise of the forgotten” by Thomas Ebel and Dr. Iris Heckmann gives an introduction on how to increase resilience in the current COVID-19 situation and beyond.
Given today’s VUCA world and increasingly intricate value chains, value chain planning has become an actual challenge, with firefighting as a day-to-day reality. Product variety has rocketed in the last decades. Market and supply options, as well as cost pressures, have risen tremendously. Products themselves and their manufacturing have become more complex, and the increasing speed of change all contribute to increased variability and uncertainty.
Several additional trends add to the challenge, e.g., more extensive, progressively globalized value chains and increasing global risks like trade wars, natural disasters, and pandemics as well as a constant push for value chain efficiency. A lack of access to talent to run the value chains of the digital age makes the situation even more demanding.
Responding to the challenges: the route to success
Adding more resources to the planning function often does not result in satisfactory outcomes as planning processes depend on tribal knowledge and manual adjustments. Why? The carefully designed plans that span value chain stages and time horizons are known to be outdated the moment they are created. Everyone knows to distrust the plans, and that an expert planner is needed to make sense of them and do the right thing.
Nonetheless, significant amounts of money are spent on complex planning systems, promising fully integrated plans by adding even more detail. “The more, the better,” the thinking goes. Experienced planners have seen it come and go. Therefore, there is vast untapped potential for productivity improvements as well as significant quality improvements, both adding directly to business value.
But how to do it? We believe there are generally two routes ahead: the first comprises deep AI approaches based on (classical) hierarchical planning. The innovation lies in applying optimization and agents to resemble what has been done by generations of planners. Except that now, AI will strongly support the human planner. The second route ahead is what we call smart parameter-driven planning. In the next section, we explain why we prefer this approach.
Parameter-driven operating models link the chain with parameters, not with plans
In a smart parameter-driven planning approach, parameters link all planning steps along the value chain and across time horizons. Parameters include lead times, buffer levels, and order priority. These parameters also serve as the baseline to monitor performance.
Using parameters, instead of continually changing plans to link the value chain, formalizes the chain’s configuration, makes it explicit, and ensures consistency across all planning stages and horizons. Accordingly, buffers calculated with harmonized rules based on parameters like actual demand, variability, and minimum order quantities, can be determined automatically and consistently throughout the whole chain. Also, simulations can reveal the impact of parameter changes in advance without the need for complex plans.
Besides, an alignment of a lower-level decision with the upper level is only required if the lower-level decision crosses the allowed boundaries. For instance, if a target range for inventory buffers is defined on the tactical level based on anticipated demand and capacity projections, then the operational level inventory decisions can be configured within this range.
The result: the parts of the value chain can operate on their own like self-regulating control loops within the precise confines of the parametrization. Apart from exceptional situations, the stages and horizons are in sync without the need for constant adjustments to the plans to keep them under control. Clear rules govern how to determine the parameters which link the stages.
Parameter-driven has been around for quite some time – now it also speaks value chain
There already exist several approaches that follow or integrate a parameter-driven idea. Lean production, for instance, contains a vast range of measures for an efficient value chain, among them Kanban as a parameter-driven pull replenishment concept. Also, the Theory of Constraints (ToC) includes many parameter-controlled aspects, like the Drum-Buffer-Rope, including its bottleneck rate concept (drum) – called takt time in Lean. The latest approach, Demand-Driven MRP (DDMRP), combines the benefits of its ancestors and uses parameters to configure a network of connected stock buffers and to drive end-to-end value chain synchronization. Markus Kuhl covers the contribution of the parameter-driven value chain approach to DDMRP in the CAMELOT DDMRP primer “Why now is the time to transform supply chain planning”.
Conclusion
We are convinced that a parameter-driven value chain is the route forward that works much better than a traditional value chain centered around hierarchical planning – be it with or without the latest technology enhancements like AI or ML. The reason for our conviction is that linking the value chain across stages and time horizons has many advantages compared to mastering an abundance of individual plans.
In the upcoming articles, we will explain in detail why and how a parameter-driven value chain works and what type of benefits you can expect. So stay tuned!
We would like to thank Sebastian Hild for his valuable contribution to this article.
