Nowadays we can do everything from grocery shopping to getting answers to complex problems and even controlling the lights and temperature in our home by simply asking smart speakers! The time has come to have the same capability in controlling our supply chains. We are now at a point to have the capability to model the supply chain accurately enough to be able to ask the system to deliver executable and optimized plans. Many, if not all, current Sales and Operations Planning S&OP systems are about 60% accurate because of the way they model the supply chain. Basically, using spreadsheet technology of fixed lead times, bucketed capacities, and pre-defined bottlenecks. Hence, much user interaction is needed in order to make the plan executable. Given the millions of different variables and objectives, it is impossible for planners to guide the system to a plan that is optimal. Often compromises are made just to make it work. With the use of Artificial Intelligence (AI) techniques we are now at a point for the system to take all such variables into account and produce optimal and executable plans with a simple instruction: “Optimize my supply chain.”
Very soon we shall be using self-driving cars on the roads. They allow you to specify the destination, they then generate a plan, monitor and execute. Based on monitoring of the road, they make adjustments in the plan to get you to your destination. While they are driving they take into account information from many different sensors (variables) in order to avoid collision or delays. An intelligent supply chain planning system behaves in the same way. First, it must have an accurate model of the environment (analogous to the street map of the GPS system). Secondly, it will create a plan that is optimal taking into account the current roadblocks and shortages etc. As it executes the plan, it monitors the execution (e.g. a machine breakdown, or late delivery by a supplier) and comes up with alternative routes until it reaches the destination of delivering the goods on time. Much like a GPS system, the planning engine is also capable of predicting potential problems in future as planning and re-planning is done.
In addition to self-executing the plans, intelligent supply chains can constantly improve their performance based on what they observe and underlying trends that are happening in the supply chain. We refer to such systems as self-improving supply chains. Many such trends may not even be visible to the users because of its subtle and slow rate of change. But Machine Learning (ML) algorithms can detect such trends and avoid the problems even before it happens. Thus, making the supply chains risk resilient.