Since its founding in 1977, the Nordic logistics consulting firm Langebæk has specialized in warehouse development, including warehouse automation. Over the years, the company has observed how warehouse operations have become increasingly data-driven—that is, reliant on data and analysis to optimize productivity, efficiency, sustainability, and customer service.
“During my twelve years at Langebæk, developments in data analysis and the use of data have perhaps been the most noticeable trend,” says Morten Badsberg, a director at Langebæk with special responsibility for developing the company’s expertise in this area.
About 80 percent of Langebæk’s projects involve assisting client companies in various ways with warehouse automation. And in these projects, it is of the utmost importance to have accurate data.
“Having and using accurate data is a prerequisite for automation. We can’t automate a warehouse unless we know what we’re handling—in terms of volumes, dimensions, and so on,” says Morten, offering an illustrative example:
– If we have data indicating that an item fits in a box measuring 60×40 cm, but it turns out to be two centimeters larger, we’ve run into a problem that can create a lot of extra work when developing an automated warehouse solution.
”I got a set of perfect data”
He cites the example of a large Nordic retail company that, following a large-scale automation initiative, delivered a perfect set of data—something Morten had never seen before.
– There were no errors in the data we received, and no data was missing—everything was perfect. The reason was simply that they had already automated their processes, which had forced them to gain full control over their data. We now see a big difference here compared to ten years ago, when it took much more effort to obtain the data we need.
Custom Data Model
Keeping track of basic data on shipments, orders, items, volumes, weights, dimensions, load carriers, packaging, and more has always been important. But with the increased use of technology in warehouses, the need to utilize data has become absolutely crucial. And for Langebæk, data collection is generally the first step in a client project. That’s why the vast majority of projects begin with the team or consultant in question clarifying with the client exactly what data is needed. They then use their own data model, based on many years of experience and best practices.
– Using our model, we can quickly assess the quality of a dataset and determine what additional data is needed. We enter fairly detailed data into the model regarding inbound and outbound flows, processed items, and so on. Based on this data, we propose an optimal solution that is either automated or entirely manual.
Varying quality and ability
The extent to which companies are skilled at collecting, analyzing, and using data varies considerably. Larger companies are generally better at this than smaller ones, but Morten emphasizes that there are significant variations.
“Larger companies have generally become quite proficient. Especially when they’ve automated their processes. If you’ve invested in a complex warehouse automation system, you need dedicated staff who focus specifically on collecting and analyzing data in order to optimize the facility,” he says, noting that it can be difficult for a smaller company to justify the cost of staff who work with data and data analysis in logistics.
More data scientists and analysts
Larger companies are now hiring dedicated data scientists and analysts to capture and utilize logistics data as the foundation for both warehouse automation and the development of automated data capture, analysis, and decision-making using AI and machine learning algorithms. Langebæk is also increasingly recruiting people with similar skills.
“The younger people we’re recruiting from universities these days often have a solid understanding of AI, machine learning, and Python. That’s good for our own internal professional development and, consequently, for our customers,” Morten notes.
—
Interview and text by Stefan Karlöf
