Jun 29, 2022
When Humans Reach Their Limit, Machines Can Take Over
Artificial intelligence
Have you ever taken questionable business decisions because you weren’t able to predict all eventualities or consider all the options available? Have you ever been stuck with an excess of inventory that could have easily been prevented? Have you ever been caught off guard by production and management challenges that were too big or too complex?
If you’ve answered “yes” to any of these questions, it might be because your planning is still done manually. We must face the music: calculations and complex problems are far from being our cup of tea. For instance, a study from Queensland University, in Australia, was carried out to know our ability to resolve problems when we’re faced with multiple variables. Results showed that above three or four variables, our capacity for analysis, calculations and data processing reaches its limit.
Luckily for us, there are simple, efficient and accessible technologies whose sole purpose is to make our life easier. After all, computers were designed to do calculations and have long proven their efficiency in this area – because where the limit of human analysis stops, that of the machine has only just begun.
A good example of planning and logistics is a method called “operations research,” an advanced statistical analysis approach that greatly simplifies decision-making, and can even make it for us. In short, operations research takes a set of data and variables and uses them to obtain an optimal solution. It meets very specific needs: with the appropriate data, logistics and a few calculations, you can benefit from an automated and precise management of your operations within previously impossible deadlines.
Let’s take, for example, a company that specializes in wooden floors. If you have hardwood floors at home, you know that the boards are usually of different lengths, ranging from one to eight feet. Nobody wants to receive dozens of very long boards and then only a few short ones. In the same vein, the company does not want to deliver boxes with only the same format of boards, and thus unbalance its inventory. Several variables are taken into account when comes the time to assemble the boxes, such as the availability and the length of the boards, the customers’ expectations or the area to be covered.
The problem is much less simple than it lets on. The goal here is to obtain a perfectly balanced mix of boards with the help of combinatorial optimization, a branch of operations research. Like a giant game of Tetris, robots are tasked with distributing the planks in random order and at a dizzying speed, so that all the boxes are well diversified. In doing so, the Company makes sure to have balanced boxes to satisfy customer needs.
There are so many uses to operations research that I can’t begin to summarize them all. From logistics to the planning of work schedules, routes, and supply chains, operations research is still largely unrecognized in relation to its real potential on many business models.
In factories, it allows you to manage inventories in real time and balance them continuously, in addition to ensuring optimal production to reach your target profit margin. These same algorithms can also be used to dispatch personnel and equipment, develop schedules and use your resources to their full capacity.
That is why when the time comes to talk logistics with local businesses, I am often surprised to see that so few of them have considered automating their decision-making processes. For many, these tasks are still entrusted to employees who, despite all their expertise, their years of experience and their willpower, spend entire days in front of gigantic Excel files and who unfortunately never manage to achieve the results that a machine can deliver in minutes.
Of course, the idea is not to replace your entire staff with computers and robots. In many ways, humans will always be essential in our businesses. But please, let's stop asking our employees to do multi-variable calculations hoping for optimal performance!
Instead, let us use operations research as a decision-making tool to free humans from complex calculations so that they can focus on the strategic aspects of their task. Ditch the endless Excel files stuffed with data of all kinds. Let's give credit where credit is due, and let the machine do what it knows best!
