Operations Research in 5 Minutes
The reason some businesses fail while others thrive.
Why should you care about Operations Research?
For a business to thrive in today’s competitive marketplace, it needs to make sound decisions consistently, and operations research is the best way to do that.
That means if you understand operations research, how it works, and the role it plays in business decisions, you’ll be more desirable for any business or strategy role.
Operations research is foundational to effective supply chain management, which is the backbone of a business. In the supply chain, businesses make decisions about production, transportation, pricing, and fulfillment.
These are exactly the types of decisions we can make with operations research:
- How should we schedule our truck drivers to transport products to minimize costs?
- How much should we produce each month to meet demand while limiting strains on inventory?
- How should we price our products?
- What should we produce? What products are hindering profitability?
…and many more.
What is Operations Research?
Operations research is at the intersection of business and mathematics.
Many decisions in business, especially in operations, are complex. Business managers can’t make these decisions based on their personal judgment, so they need to rely on mathematical models.
In operations research, we use optimization, statistics, and computing to build mathematical models. This allows us to take a real-world business problem, convert it into numbers, and put it into the model to determine the best course of action.
The Operations Research Process
I illustrated the operations research framework for you below:
Starting with a real-world problem, we formulate it into a mathematical model, then use computational tools to solve the problem using the model to get a numerical solution that represents the optimal decision. At the end, we interpret the business meaning of that optimal solution by converting it back to our real-world context and implementing the associated decision.
The purple boxes contain results, and the arrows show what a researcher does to get to them — formulate, solve, and interpret.
These are the 3 core components to understanding the field, so we’ll briefly expand on each.
Formulate, Solve, and Interpret
1 — Formulate: from Problem to Model
There are many types of formulations, and the most used ones are Linear Programs, Integer Programs, and Network Flow Problems. You don’t need to know the details for a high-level understanding, but you should know the components they’re all made up of.
3 Components of a Formulation
- Variables: convert decision parameters into mathematical variables that will be inputted into the model.
- Objective function: define what we’re trying to maximize or minimize ie. maximize profits or minimize costs.
- Constraints: the solution must satisfy a set of mathematical constraints.
Constraints model the restrictions in the real world. For example, we typically have a limit on how much our machines can produce or how much raw materials we can use.
Without constraints, no decision is needed — business decisions are so hard only because there are constraints. We can’t have it all, so we must make a trade off.
2 — Solve: from Model to Solution
We solve the model computationally to arrive at an optimal solution, within our constraints, that either maximizes or minimizes what we wanted to optimize for, whether it’s revenues or costs.
This can be done by hand in a process called Simplex, but it’s built into computational programs and easily solvable in any programming language (like Python). If you’re interested in the technical side of things, I recommend checking out Samir Saci, he gives great explanations.
It’s often a good idea to do sensitivity analysis to test the robustness of the solution. This is when you alter certain parameters to see how big of an impact they have on your final decision in case your assumptions are off.
3 — Interpret
This is the final stage. We have an optimal numerical solution, and we need to convert it into a tangible decision.
It goes back to understanding what each of the variables means contextually, so we can interpret them correctly as implementable production, scheduling, distribution, or other decisions.
We then implement this optimal decision in our operations.
Congratulations! You now have a big-picture understanding of operations research — I hope this knowledge will help you in whatever you set out to do.
Here’s a quick recap of what we talked about:
Thanks for reading.
Follow me so I can bring more quality content to you. Also, you might benefit from my other business articles:
