The notion of continuous planning has gained some popularity recently. It is a departure from batch type of planning on a monthly, weekly or even daily basis. Continuous planning is intended to make sure that the plans are not outdated since most batch type of planning may take a few hours; and by the time it is available the conditions may have changed resulting in an outdated plan. But the question is why continuous? There are more effective methods of keeping the plan up-to-date. One approach is impact-driven planning, also known as event-driven planning.
Consider an analogy, driving from LA to SF. If everything is going according to the plan, with some minor deviations, then why do we need to replan or do continuous planning unless you see an event or predict an event, such as a road block ahead, that may potentially have an impact. To this end, understanding the impact of an event and re-planning accordingly is a better approach. We call this impact-driven planning. That is, based on the importance and the impact it has on the current plan that decides the need for re-planning. Examples are a large unexpected rush order from and important customer, supplier is late causing idle equipment and late customer deliveries, or bottleneck equipment break down amongst others.
In order to be able to perform Impact-Driven Planning (IDP), intelligence is needed to interpret the impact of the event. For example, not every late supply delivery causes missing customer due dates. Moreover, there needs to be an understanding of the cost/benefit ratios and tolerances of customers, KPIs and management to evaluate the true nature of the impact. For example, a customer may tolerate a few days late but not a week. A production manager may decide to get some orders done earlier to free up capacity later at the cost of higher inventory cost.
Furthermore, a decision needs to be made if the event requires immediate re-planning or it can be done later because we have time. For example, the impacted order is needed far into the future.
Essentially, the process requires an evaluation of how much of an impact an event would have on the existing plan. A bump on the road does not require changing your route. A major traffic jam down the road does. A stormy weather forecast from a critical supplier location may need immediate attention.
There are also different levels of reaction: bumps on the road are taken care of by shock absorbers, changing the route is performed by GPS constantly evaluating the arrival time deviations. The system needs to be layered to respond/predict to these events in a timely manner.
Lastly, if there is a change in plan, who needs to know? GPS can change plan without telling the shock absorbers, but it may need to send the change in arrival time to the destination, or may prompt re-charging or re-fueling the car.
Different stages of IDP are summarized below:
- Detect the Event
- Measure the Impact to Decide what change, if any, is needed
- Respond if so needed
- Correct and inform as needed. In other words, who is impacted
At Adexa, we use a multi-modal approach to interpret the impact of the events by performing heuristic analysis, measuring tolerance of KPIs and performing, as needed, quick automated scenario analysis. If it is decided that the impact is high and beyond certain tolerances of some KPIs, then re-planning is done either partially or fully as needed. An example is arrival of a rush order. The expectation is to return immediate real-time response for a commit date based on ATP/CTP. This is a very complex operation that requires real-time data on previous commitments, availability of materials and capacity and the level of customer’s desire to have it sooner than later and its importance. Using the heuristics and analysis described earlier, very accurate commit dates are given without having to re-plan. Even though re-planning maybe needed but the analysis concludes that the delivery can be made as promised without having to immediately re-plan.
Impact-Driven Planning is made possible by the use of AI heuristics and embedded knowledge as well as business rules specific to the company. It gets better the more it is used and learns over time how to better respond to events and measuring their impact. Needless to say, that using IDP, you no longer need to receive hundreds of alerts every day warning about things that do not really matter. And that can make your life a lot easier and more productive. For more information on IDP click Here.