The goal of production scheduling is to use people and equipment as effectively as possible to maximize overall value for the business. This means determining how much of each product to make based on costs and market conditions, and then deploying the necessary resources to get it done. Production scheduling is always a complex calculus with many decision points, dependencies, and constraints to balance.
Production scheduling is a need across all manufacturing and production businesses, regardless of industry. This tutorial is about industrial baking, but the solutions you’ll build in this tutorial could be translated equally well to any other use case that involves people and equipment making products.
Read more about the case study and the agents that solved it in our whitepaper Use Cases for Intelligent Agents: Production Scheduling.
The production scheduling case study is set in an industrial bakery that makes cakes, cupcakes, and cookies every day. Even a small production facility requires a staggering number of decisions of how to assign workers, equipment, and products over the course of a day, with each set of decisions potentially leading to different results. To make products at scale, the complexity only increases.
In addition to this complexity on the production side, production scheduling use cases also have to contend with variable external factors, including demand, product pricing, cost of materials and labor, storage and transport factors, and potential penalties for failing to meet demand.
These are the kinds of considerations that are difficult for humans and traditional optimization technologies to solve because the variables and their relationships are tremendously complex.
That’s why agents engineered with Machine Teaching are so valuable for production scheduling use cases.
The top-level goal for the bakery is to maximize ROI – that is, make as much profit as possible by increasing revenues and minimizing costs. To do this, it needs to produce the right amount of each of its products, using the workers and equipment it has available to meet demand and maximize profit.
But what does that look like? There are many different strategies that might be used to try to reach this goal. No one strategy is perfect, and different approaches might work better at different times depending on the conditions.
