In this article, we provide two scenario planning examples that demonstrate why running prescriptive analysis on top of a digital planning twin (similar to a supply chain digital twin but more representative of the actual business) is the only way companies can ensure optimal performance during and after the COVID-19 pandemic.
Supply chain leaders need to be able to set constraints and adjust assumptions — in real time — based on things like plant closures, crisis management and prevention, sourcing issues, labor availability, and more. But they also need to understand the forward-looking impact of their decisions on performance metrics across their value chain, such as:
- Service levels
- Production costs
- Transportation costs
- Financial KPIs (profit, NPV, cash flow)
- Sustainability requirements
- And more…
Review of the Impact of Coronavirus on Global Supply Chains
Before proceeding, let’s consider some of the ways that coronavirus has already impacted global supply chains:
- Vehicle production has been disrupted: Most automotive manufacturers have experienced supply shortages and been forced to cut or limit production. In several instances, factories have been idle for weeks at a time.
- Demand has plummeted: Demand for commodities has fallen to exceptionally low levels, and oil is at prices last seen in 2002.
- Airline travel has collapsed: Airlines are flying empty or with cargo and many are stopping international flights.
- Industry shutdown: Industries and companies in many countries, including the U.S., are being forced to close.
As dramatic as the immediate impacts are, we also need to prepare for the aftermath. It’s quite likely many businesses will collapse and that it will take months, if not years, for world economies to recover.
Global Supply Chain Planning Strategies Need to Change
That the world was unprepared for something like the coronavirus pandemic is obvious. But when you look back over the last century, it was surely only a matter of time before a pandemic like this was going to happen again. Even a cursory application of historical and predictive analytics would have pointed to this possibility.
This points to an unfortunate weakness in current supply chain scenario planning processes and illustrates the lack of robustness and resilience to natural and other globally disruptive disasters. Perhaps it’s because everyone has been laser-focused on reducing costs and has thus pushed aside or deprioritized the costs associated with building robust, resilient supply chains that are still able to perform optimally in the face of disaster. In the fast-paced markets across the supply chain, it’s understandable that companies have this mentality. However, if coronavirus has taught us one thing, it’s that now’s the time to invest in technology, strategies, and processes that will ensure optimal financial performance in times of heavy disruption.
Take sourcing, for example. Those who rely heavily on global sourcing, especially from areas hit hard by the coronavirus, have seen tremendous disruption. Even worse, many who rely heavily on global sources didn’t have preventative plans in place that would minimize the impact of losing one or more key suppliers. Compounding the situation is that, despite early warning signs of major disruptions, few if any companies knew what corrective action to take. And now, the global supply chain industry is in rapid response mode, forced to make decisions that aren’t data-backed. Even worse, the forward-looking impact of these decisions are unknown, for the most part, and threaten to remain unknown as operations begin to restart.
Why Prescriptive Analytics for COVID-19 and Supply Chain?
Prescriptive analytics is known for being the only form of analytics that can recommend an action based on any number of constraints and one or more objectives (key performance indicators, including things like profit). Thus, unlike using forecasting and educated guesses to determine the best plans of action during the COVID-19 disruption, prescriptive insights would output the best decision. As much as possible, prescriptive analytics removes human bias from decisions while letting organizations clearly see the forward-looking impact that every decision would have on every aspect of the business.
The other crucial component we mention above is a digital twin of the business. A digital twin is a digital representation of some real-world entity or system, usually a representation of assets and processes within or across a business. The concept of a supply chain digital twin is nothing new, and numerous technology vendors offer digital twin capabilities that can be used to optimize the use of assets, like when to run upgrades, how to go about repairing machines, etc. A supply chain digital twin effectively allows companies to run simulations and understand the impact of changing one or more aspects (e.g., adding or removing a manufacturing plant) on supply chain KPIs.
There are a few different kinds of digital twins when it comes to supply chains. Some only represent one aspect of the business while others represent the end-to-end value chain. COVID-19 has shown us that for companies wanting to ensure financial success during times of significant disruption, a supply chain digital twin isn’t sufficient. Instead, companies must leverage a digital twin that includes parts of the business beyond just the supply chain network — including financial assumptions and implications. At River Logic, we and our Partners refer to this as a digital planning twin.
By leveraging a digital planning twin, planners can run prescriptive analyses to create forward-looking short-term (days/weeks), mid-term (months), and long-term plans that optimize assets (labor, machines, product availability) while minimizing the impact on costs, revenue, service level, and profitability to tackle any number of future situations.
Here are two real-world examples of how companies could have leveraged scenario planning, powered by prescriptive analytics, to better plan for major disruptions during the coronavirus outbreak. The goal is to create forward-looking plans that supply chains could execute, thus ensuring their supply chain remains robust and performs as optimally as possible.
War-Gaming Constraint-Based Optimization
One of the fundamental problems with conventional supply chain planning engines is their inability to identify an optimal strategy. Granted, they will inform where and what and how, or in other words, what is happening in your current supply chain, but nothing else. At best, they can operate in predictive mode; at worst, they simply reflect the current situation.
Additionally, they are unable to assess or predict the effects that making supply chain changes will have on the business and its environment.
Using prescriptive analytics (optimization) and a digital twin changes that. Because the digital twin represents how the real business functions, it’s possible to evaluate how different scenarios impact the business and identify optimal supply chain decisions. The prescriptive component (i.e., linear programming), is then able to respect and take into account any number of constraints and objectives within and across each component of the digital planning twin.
In a situation like COVID-19, teams could play war games with this technology, devising alternative scenarios and running what-if analyses through the model to determine the best course of action to take if supply chain disruption occurs. In this way, you can avoid more predictable catastrophic mistakes and understand how to manage different risks. Naturally, companies cannot plan for every possible risk. However, even just having this type of technology in place when disruption occurs means companies are much better armed to execute in such a manner that aligns with overarching business goals.
How Many Power Cord Suppliers Do You Need?
Have you ever stopped to think about where your power cords come from? In all likelihood, they come from China; hardly anyone else makes them. If you’re only making one or two items, that’s not an issue, as you can always arrange an alternative supplier, but when you’re making millions, the lack of the humble power cord can bring your factory to a screeching halt.
In a similar vein, Apple faced numerous supply problems because of its almost total reliance on just-in-time (JIT) supply from Chinese factories.
The question is how many power cord or other componentry suppliers do you need? The traditional approach is for three suppliers, but this means only one will offer the best price. You are actually paying to keep the second and third supplier in business. Being such a common item, you could have only one supplier and a list of 30 or so alternative suppliers you can turn to. Unfortunately, you’ll then have to negotiate with them if your supplier fails to deliver; the components may cost, say, 20% more, and there will be a delay before you receive supplies. This is what happened to Apple — their suppliers failed, and they were stuck making decisions with very little forward-looking data to guide them to the right decision.
During times like a pandemic, suppliers could become strapped for cash (or shut down altogether), so they may ask for different terms that pay them sooner. Others might offer higher rates but longer terms.
So, what’s the best approach? This is a typical example of a what-if analysis you can conduct using prescriptive analyses on a digital planning twin. Which supplier do you go with? Do you bring one on and then another? What is the financial impact if you’re trying to meet as much demand as possible but still conserve cash? These questions are not always as straightforward as they seem, especially as you begin considering the complexity of your end-to-end value chain (like transportation constraints, etc.) and dozens of products.
We Need More Robust Supply Chain Analytics to Deal with Disruption…Now
The simple scenario planning example outlined above illustrates why even the most seemingly simple situations can be detrimental to company-wide performance if there is no supply chain resilience built in.
We’ve all seen that it’s crucial to have fast scenario planning capabilities to accommodate changes when things go wrong. However, simply having these capabilities isn’t enough to keep supply chains performing optimally with respect to the financial goals of an organization, at large. Scenario analyses need to look beyond supply chain, and they need to account for financial goals and constraints, as well.
In order to build a resilient supply chain, companies need to begin war-gaming what-if analyses on a digital planning twin that can output optimal plans given any number of hypothetical situations. For unexpected situations, the scenario analysis technology already exists, and robust risk mitigation and intervention plans can be quickly put into place. All with the understanding that decisions are truly optimal — and not just for the supply chain.
Although this pandemic will end, it’s important to bear in mind that at some point in time another type of coronavirus will appear. Or if not a pandemic, some other natural or manmade disaster will occur that will disrupt your supply lines. But if we learn from the COVID-19 pandemic and set our supply chains up for resilience, we’ll be in a much better position globally to deal with more such disasters.
The coronavirus pandemic has brought to light the need for rapid and robust scenario planning processes, and the need for optionality in our decision making. Building resilient supply chains so as to avoid the massive interruptions and disruptions we’re seeing from COVID-19 means understanding all decisions options, being agile and flexible in decision-making, and knowing the future impacts of decisions as best as possible.