Bridging the gap between S&OP and autonomous planning

Dr. Cyrus Hadavi is its CEO Adexaa leading AI-powered supply chain planning company with Fortune-class clients on five continents.

Many, if not most, companies that implement S&OP (sales and operations planning) solutions have benefited from the long-term visibility they provide. However, they face some real problems with executing plans in the short term, whether it’s today or sometime in the next three months.

As pointed out in a recent Oliver Wight paper, the symptoms are many. Some examples are that plan changes are not reflected in the S&OP process in a timely manner, crises and ad hoc meetings are frequent, frustrations between planning and execution people, imprecise commitment dates and a rush to expedite and unachievable plans due to incorrect assumptions in the S&OP model. The chaos continues simply because S&OP is a rough plan based on high-level assumptions. Things are constantly changing, and unless the system is able to understand each event and has the ability to model their causes as well as their impact, then this is another case of segregated silos (ie a ‘vertical silo’ of design and implementation).

S&OP assumes that materials have a fixed lead time and will arrive on time. If not, then you fix it. S&OP assumes that production lead times remain the same. This is false. They depend on product mix and changing bottlenecks. S&OP assumes that resources are always operating at, say, 80% or even 100%! Not true—they change depending on many factors.

To this end, S&OP makes many assumptions that are not true. This leads to unreliable commitment dates as well as inaccurate financial forecasts, as events during execution such as the use of more expensive substitute materials, delays or changes in production are not reflected in the S&OP. In addition, we have the issue of materials not arriving on time or surprise high-priority orders that may require someone else’s immediate advance. S&OP’s remedy to this problem is to add more false assumptions by adjusting bin capacities to allow more slack for contingencies, not knowing if this strategy causes underutilization of resources, simply adds to raw material and WIP inventory, or potentially delays existing orders.

To address these issues, there is a need for an integrated planning and execution environment as a continuum—not separate processes! Disjoint S&OP and S&OE solutions develop different models and different assumptions. Two different systems that have separate assumptions cannot provide a smooth flow of operations, and this leads to simply “automating” an inferior process and practice.

A human brain plans and then constantly receives feedback while executing the plan. If you decide to eat, the brain plans to have food (what, where and how to get it). As you perform (i.e. walk to the fridge, open the door, look with your eyes, pick up what you need and put the food in your mouth), then each step is monitored and orchestrated with the brain in real time until the task is done . Imagine doing that with separate processes? Your brain thinks about having food now, but you can run it eight hours later. By the time your brain is informed that you have skipped eating, it may decide to ask for more food, but then the food may no longer be available. Thus, hunger persists.

Having S&OP and S&OE in a unified environment allows operations to be synchronized while planning and executing in real time. Accurate assumptions are made leading to accurate and reliable commitment dates and financials. Trends and deviations are monitored and “learned” by the system so that corrective action is taken using self-correcting and self-improving algorithms. By having a continuum of planning and execution, the antiquated processes of the 80s can be transformed into a much more flexible process of planning and executing on an ongoing basis. Integrating the two into one has many advantages, including efficiency, faster decision-making and, of course, much lower operating costs and much higher levels of customer service.

Perhaps you’ve already implemented an S&OP solution and are struggling with some of the accuracy issues and manual adjustments that are inevitable to make the plans work. At this point, having an S&OE solution (also known as a response programming solution) that can truly represent a digital twin can help you have much better visibility into the execution of your plans.

This does not mean that you have to immediately provide detailed data. It implies that an S&OE solution is capable of adding a better and more comprehensive modeling capability to make S&OP plans more reliable and accurate. The goal is to reduce the time to plan and implement the plan and provide accurate feedback on the S&OP process to make the right assumptions. It will also assist with all performance surprises (eg late arrival of materials, surprise orders, equipment breakdowns and unforeseen disruptions due to weather, floods, invoices and order cancellations). It may not be the optimal strategy, but it bridges the gap.

On the other hand, if you want to implement an S&OP solution, you need to ask the following questions: How can I execute the plans? How good are the programs created by S&OP? And how much time should I spend to make them executable? Ensure that the system has sufficient modeling capability (creating a true digital twin) and the ability to predict and respond to real-world events such as snowstorms, equipment failures, truck delays, floods, labor shortages or commodity shortages, and price changes . This can only be done in an environment that is capable of having both S&OP and S&OE on a continuum and modeling the true capability of resources—not bucket capacities.

This is your path to an autonomous supply chain. Be the first in your industry to lead and break out of processes designed around the spreadsheet capabilities that came on the scene in the 80s. The digital age demands more – a design transformation.

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Am I eligible?