2020 was a very challenging and difficult year for many supply chain leaders. Climate, tariffs and the pandemic unraveled how fragile supply chains can be. Needless to say, that such disruptions are not uncommon and happen frequently resulting in hundreds of millions of dollars of losses due to climate, geopolitical factors, Acts of God, fire, and ethical issues such as child labor.
2021 will not be an exception, many industries may face a tsunami of demand over the next few months and some may face yet another episode of low demand and potential loss of suppliers, loss of revenue or even bankruptcy. We have already witnessed quite a few in the retail industry as well as others such as entertainment, restaurants and hotels.
Resiliency of supply chain is the ability to withstand changes in demand and supply and the ability to respond effectively. The short to medium term response could be changes in product volume or mix, add or remove suppliers, or ensuring cash availability for survival and recovery. Having a digital supply chain can help to respond faster to these disruptions. However, having the right design of the supply chain minimizes the risk and effort needed to respond to changes that are out of our control. Prediction is a much more effective tool than response. Both are needed; however, the more you can predict the less you need to respond. To this end, having AI and ML tools to make predictions and prepare for potential issues in advance is the right thing to do. Prediction comes in different forms. It can be used in the design of the supply chain. In other words, making the supply chain as “parallel” as possible with redundancies built into it. This is obviously more costly however it is the insurance premium you pay for safety of your supply chain. Prediction can also come from observations of past patterns. Not all disruptions are random such as pandemics and earthquakes. Although one can design for earthquakes but predicting them is difficult. ML is used now a days to predict certain events that are more likely to happen during a certain month of the year. These are factors like supplier performance, changes in demand patterns, pricing of commodities or hedging and supplier variability due to geopolitical factor or other causes. Using such predictions, it is then possible to use prescriptive algorithms to ensure uninterruptable operation of the supply chain. The latter is done by automatic adjustment of inventory levels using ML or by increasing or reducing production or supplies or even activating additional sources for manufacturing or sourcing. Not all disruptions are predictable but most are and using a combination of supply chain design and AI/ML tools the supply chain can become quite resilient. For more on this topic and use of AI/ML for risk mitigation click Here.