If one is a tennis fan, they probably watched the finals last weekend. What was interesting was that IBM Watson gave the loser of the women’s final almost twice as big a chance to win than the actual winner of the game. The question arises, what if this prediction approach was applied to inventory values? It would result in almost 30% more than what would be needed. People are all too enamored with AI, machine learning, and more recently generative AI, but forgetting that these techniques have limitations and all they do is look in the rearview mirror to decide what is ahead. What lies ahead can be very different, as experienced with the pandemic.
Obviously, in many cases human intelligence can help to do better predictions. However, that is also skewed by one’s own experiences and biases. There is nothing wrong with using predictive tools to see what might be ahead and use of past data. However, more can be done to reduce risk and increase resiliency. Resiliency and agility are really the key to overcoming disruptions and benefitting from unexpected opportunities.
Resiliency implies having the ability to mitigate risk and be able to respond to potential issues or opportunities. To this end, one needs to know what the supply chain is capable of, so that as soon as an issue comes up, the system can quickly examine all the alternatives and make a recommendation. Therefore, one needs to have an accurate and up-to-date digital twin of the supply chain. Note that this approach is different from manually examining all the different scenarios. Manual scenario planning is time consuming and can only look at a handful of scenarios that the planner might be aware of. There are thousands more that should be examined and can be examined in seconds. We can never expect perfect forecasts and perfect levels of inventory. However, we can expect the best answers from the system by having the ability to make decisions based on prediction and what the best options are every time a disruption occurs. Best option means the optimal combination of objectives set by the users including cost, service level, environmental factors, revenue/profit, market share and more.
For more information on how a continuum of S&OP and S&OE can enable resilience and agility click Here.