A planner once asked me whether our planning automation solvers are programmed to be like cats or like foxes. When I stared back in bewilderment, she explained that she was referring to the Aesop fable in which the fox boasts about having multiple tricks and dodges to escape from enemies. The cat confides that she has only one, and when hunters arrive with their dogs, she quickly climbs a tree. The fox, pondering which among the wide range of strategies to go for, is slow to react and is caught by the hounds.
Is it better to be a cat then? That’s not what we would say here at OMP. Solvers are created to be smart, versatile, and flexible, just like the fox in the fable. Solvers can be fine-tuned and adjusted to take into account operational and organizational constraints, optimizing any plan to provide maximum service and profitability. All kinds of conflicts and emergent issues can be reckoned with, including stockouts, machine overloads, expiration or obsolescence, inadequate campaign durations or changeover frequencies, excessive waste, lost sales, and customer dissatisfaction. In general, solvers are also nimble enough to avoid catastrophe and ‘escape the hounds,’ despite the fact that balancing conflicting constraints can take up a lot of computing time.
Advocates of the ‘touchless planning’ hype will argue that new generations of solvers make a more powerful case for the fox strategy. More and more elements are being integrated into the solutions, machine learning capability is further improving outcomes, and computing speed is set to skyrocket.
That’s true, but many planners still have second thoughts about relying on solvers, and the reason is not related to speed but to quality and above all transparency. Planners find that solvers are like black boxes. Have all the potential issues been taken into account? How does a solver prioritize when balancing conflicts? What assumptions have been made? Why did the solver prefer option A over option B? Planners need a sense of control and find it difficult to understand how solvers make decisions.
Supply chain planners are constantly at crossroads where crucial business decisions need to be made. Do you disappoint an established customer by seizing a promising opportunity? Or increase sales volumes only to see margins go down? Or maybe improve on-time-in-full at the cost of generating more waste? Planners often feel torn between two choices.
That’s where business rules come in. Essential additions to the planning automation toolbox, they provide guidance for planners on how to react in a given situation. Often taking the form of decision trees, they might also use models to classify customers or rank issues and risks based on KPIs and periodic evaluations at management level.
I discuss the types of business rules in another blog post, but one thing is sure: they need to be clear, unambiguous, and fully agreed across the entire company. They’re not just useful to guide planners, they also allow them to evaluate the outcomes of automated planning solvers and scenarios. And they’re a great way to bring some common sense into the discussion. Which aligns perfectly with my favorite interpretation of that cat-and-fox fable: common sense is always worth more than cunning.
Want to learn more about how to bring common sense into your planning?
BiographyMattias has worked with supply chain technology for more than a decade. He has always been interested in the art of user onboarding and has valuable experience in the requirements for proper tool adoption.