One of the explicit goals of Industry 4.0 in the long run is to empower autonomous machine decision making within production processes. This is a lofty goal—requiring highly visible and highly legible data streams combined with AI or machine learning integration—but it does have the potential to add considerable value to supply chain management processes. How does it do so? By freeing up human decision making capacity for larger-scale choices, and by automating the process by which data is turned into action—i.e. creating an implicit set of procedures for different situations that might emerge on the factory floor. In this way, manufacturers can build new efficiencies into their existing processes and drive towards an increasingly optimized supply chain.
Autonomous machine decision making might seem like a radical change for the industry, but in reality it’s just another form of supply chain decentralization. Decisions that used to be made by a centralized group of executives can now, with IoT (internet of things) integration and other technology, be taken over by decision makers (in this case non-human ones) further down the value stream. Obviously, however, it’s not the only example. Decentralization can also take the form of spreading out decision-making abilities among human planners in disparate plants or locations. This might be less exciting from a technological perspective, but it can be just as value additive.
Who Produces What, When?
On some level, the basic question in supply chain planning boils down to who should produce (and later transport) what product at what time. In a multi-national corporation with a centralized planning flow, a team of planners in the same geographic location makes that determination for every other touchpoint on the value chain. This can have its advantages: the left hand always knows what the right is doing, and for upper level strategic decisions it can be helpful to have an overarching, company-wide vision for different plants and operators to latch onto. The issues that crop up tend to be a lack of flexibility, because other touchpoints on the supply chain are always waiting for approval from a team that might have to prioritize one thing over another, and sometimes lack direct access to mission critical data (which might be most easily accessible from the various different plants).
Unsurprisingly, a decentralized supply chain is basically the opposite. Within the confines of some overarching strategic plan, lower-level decision makers are left to determine what they should produce (or transport) at what time. It’s easy to see the ways in which this could add value: more rapid responses to parts shortages, sudden demand shifts, fleet breakdowns, and other supply chain disruptions that a geographically distant team might not be in a position to respond to optimally; better integration of emerging production data into future plans; more direct contact between empowered decision makers and suppliers/customers, potentially leading to stronger relationships with both; and more active cooperation between teams. At the same time, however, it’s easy to see the ways that such a strategy could add risk. Not only does the probability of operational disconnect (in which disparate teams are working at cross-purposes) increase, you also run the risk of creating data and decision making silos—both of which can produce disruptions down the line.
How to Avoid Silos
Okay, let’s say you see the value in a decentralized approach to supply chain planning, and you’re wondering what you can do to mitigate the relevant risks, i.e. increased silos and decreased organizational cohesion. This is where additional Industry 4.0 concepts come into play. No, you don’t need to implement autonomous machine decision making in order to gain value from decentralization, but you do need to adopt a mindset and, more crucially, an IT infrastructure that promotes end-to-end (E2E) visibility. What does this mean in practice? Essentially that whatever operational data is being collected throughout your organization needs to be readily accessible across the value chain. This effectively prevents silos, meaning that you’re unlikely to find yourself in a position where a production facility in one region can’t figure out what goods to produce in what ratios because they have no insight into the production flows of any other facilities.
To take this one step further, you might even strive for broader supply chain integration—meaning that your suppliers and distributers would also have value-additive access to your data and vice-versa. In this way, you can decentralize your decision making processes with a greater degree of confidence in the ultimate outcomes. If everyone is sharing and analyzing the same data, they will largely come to a similar set of conclusions, meaning that lower level decisions could be delegated without an undue amount of risk.
Creating a Hybrid Approach
Of course, no supply chain functions in a completely decentralized way—just as no supply chain can operate with centralized teams making literally every single decision. Some overarching, centralized planning will have to be done, just as some decisions will always have to be delegated to those downstream in the value chain. This means that, while many supply chains could benefit from decentralization, every organization will fundamentally wind up using a hybrid approach. The question then becomes, how do you create a balance between centralized and decentralized supply chain management that will provide the ideal mix of acceptable risk levels and optimal value added?
Different businesses will have different answers to this question, but generally speaking, the more data you have and the more successfully you make it available across your supply chain, the more effectively you can decentralize. If we turn our attention towards relatively Industry 4.0-enabled environments, we can picture businesses adopting S&OE workflows, in which low-level planners monitor live demand and transport data in order to make daily and weekly adjustments to shipping flows and inventory levels. In this way, small, short-term decisions are leveraged to maintain the viability of longer term (read: quarterly or annual) strategic and operational plans. Here, we see a valuable balance being struck: relatively autonomous decision-makers, armed with usable data, operate within a clearly defined purview with the express aim of integrating real-time supply chain operations with the expectations of a central planning team. This isn’t pure decentralization, but something even more powerful: synergy across the entire planning chain.
