'The sexiest job of the century'

A growing team of data scientists – in what the Harvard Business Review has called ”the sexiest job of the century” – is using Maersk Line’s abundance of data to chip away at key operational and commercial challenges, not to mention to explore ideas for new business models.

data-scientists-maerk-line-1
Klaus Holst, Nicki Bolbroe and Jan Voetmann and Brian Rysz are some of the people at Maersk Line involved in using advanced analytics to transform the shipping business.

What is a data scientist?

Definitions of what a ‘data scientist’ does are long, but all of them describe a person able to leverage multiple disciplines – economics, coding, ­statistics, mathematical modelling – in order to extract knowledge and insight from large and various data sets that may be unstructured and incomplete.

More than 4 million times a year in one thousand different locations, a repair is carried out on a Maersk Line container. It costs the company more than USD 300 million, annually.

Maersk Line has recorded the details of all of these repairs going back for years – what was done, what it cost, where and when it took place – in spreadsheets. As in many other areas of the business, the data was there; what Brian Rysz needed was insight into the data.

As the Head of Maersk Line’s Global Equipment Management and Repair, Rysz wanted to know if and how the data could help him shave millions from the annual repair budget.

“Every day, tens of thousands of containers are flowing in and out of repair shops all over the world. There are so many variables to consider, i.e. the type of container, the manufacturer, the cargo that was in it, the parts used, the repair done and the location,” says Rysz. “If we could use that info to identify where we’re getting good and bad repairs from and why, we could start making a dent in this cost.”

Making the data tell a story
Back in January, Rysz was introduced to Nicki Bolbroe and Klaus Holst. Both are analysts with Maersk Line’s Advanced Analytics team, a relatively new team in Maersk Line comprising eight data analysts, experts in combining mathematics and other disciplines with computer programming to extract meaning from large and different data sets.

data-scientists-maerk-line-3
Maersk Line’s Advanced Analytics team is a relatively new team in Maersk Line comprising eight data analysts who are experts in combining mathematics and other disciplines with computer programming to extract meaning from large and different data sets.

After some initial meetings with Rysz to determine exactly what he wanted, Bolbroe and Holst holed up in a conference room for one week with their laptops and several colleagues from Maersk Line IT’s Agile Data Lab. The ADL team are data hunters, they had collected and cleaned the mountain of data the analysts needed. It consisted of repair data going back five years along with data for several other related actions carried out on every container, including ‘pre-trip inspections’ for reefer containers, and every movement and every booking over the period for each of the 2.7 million containers in the Maersk Line fleet. All in all, it amounted to several billion lines of data.

“The business tells us what they want and we have to figure out how to get there,” says Nicki Bolbroe. “We crunch the numbers, add information, make and test some algorithms, code some things, share it with the business and subsequently repeat the process until we all get to where we want to be.”

The big picture on a small screen
After a week of non-stop work, Rysz says the analysts nailed it. Suddenly, the Maersk Line world of repairs was on his computer screen in an easy to understand red and blue quilt of different sized squares.

Rysz could simply click around, country by country, city by city and “visit” each individual repair shop, and thereby see precisely where he was getting good and bad repairs. He could even sort the data by type of repair, time frame and more. For the first time, he had an overview of Maersk Line’s global repair activity; he had the insight into his costs he was looking for.

At the moment we’re moving from core business intelligence reporting to exploratory data analysis and our focus is primarily operational and commercial optimisation. In ten years though, maybe sooner, when the data quality and capabilities are mature, Maersk Line’s analytics functions will help directly grow the company’s bottom line.

JAN VOETMANN, HEAD OF ADVANCED ANALYTICS IN MAERSK LINE

Maersk Line’s Advanced Analytics team
Maersk Line’s Advanced Analytics team has a long list of projects it is working on with different parts of the Maersk Line business and many more lined up.

  • Operational optimisation and commercial excellence are the primary focus of the team
  • Besides container repairs, the team is involved in a variety of operational efficiency projects. Two of the biggest are bunker optimisation and empty container forecasting. They are multi-billion dollar cost areas with the kind of complexity the team thrives on
  • New business models are a third area of endeavour that reflects the likelihood that Maersk Line’s globe-spanning operational and commercial data – combined with the ability to analyse it – will one day be a valuable product itself

One immediate finding was that the importation of certain spare parts into India – a huge repair location – was not necessary. The data proved that locally sourced floorboard materials were not leading to a greater frequency of repeat repairs as suspected, but that the cargo being loaded there – marble and concrete – is a factor. This knowledge alone will save Maersk Line USD 2 million a year.

“This is incredibly exciting, but it’s also new. We cannot say at this point that ‘this visibility will save us this much’, but we’re already sharing the findings with suppliers to help them improve. We know it will have an impact,” says Rysz.

The sexiest job of the century
A visit to the Advanced Analytics team on a Friday afternoon confirms some expectations – complex ­mathematical equations on one person’s screen and what appears to be programming code on another. Bon Jovi’s “Wanted, Dead or Alive” plays out loud on team leader Jan Voetmann’s phone. “It’s Friday,” he says.

The AA team is comprised of eight people with varied academic backgrounds and experience, along with a lot of skill overlap in mathematics and a statistical computer programming language called ‘R’. All eight of them are data scientists – what the Harvard Business Review has called “the sexiest job of the 21st century.”

At 43-years of age, Voetmann is the Head of Advanced Analytics and the oldest in the department. He has a Masters in Mathematical Economics. He wrote his thesis about one of his heroes, John Nash, the mathematician and professor portrayed in the movie, “A Beautiful Mind.”

Voetmann’s first hire two years ago was Nicki Bolbroe, who at 27 is the youngest member of the team. He has a Masters in Management Science and specialises in econometrics. Klaus Holst is one of the latest additions to the team. He has a PhD in Statistics with particular expertise in two branches, survival analysis and event history analysis.

“More than anything, what data scientists do is make discoveries while swimming in data,” writes the Harvard Business Review in a story titled Data Scientist: The Sexiest Job of the 21st Century. “In a competitive landscape where challenges keep changing and data never stops flowing, data scientists help decision makers shift from ad hoc analysis to an ongoing conversation with data.”

“What kind of business is Maersk Line? It’s a ­shipping business, but it’s also very much a logistics business that depends on optimisation,” says Jakob Stausholm, Maersk Line’s Chief Transformation Officer. “If we want real cost leadership and commercial excellence, applying the technology and the competencies of advanced analysts can take us much further than conventional analysis.”

The analytics community is also excited to see Maersk entering this space. A conference for users of ‘R’ hosted by Voetmann’s team this past summer filled the large auditorium at Maersk HQ in Copenhagen with curious peers eager to hear what Maersk Line was doing with analytics.

“Liner shipping has many interesting problems for people like us; complex problems that need to be solved, mathematically. It’s very challenging and motivating. It’s an exciting time to be in this field,” says Holst.

The future of shipping

The AA team has a long list of projects it is working on with different parts of the Maersk Line business and many more in the queue. Operational optimisation and commercial excellence are the primary focus of the team and these projects dominate the list.

Besides container repairs, they are involved in a variety of operational efficiency projects. Two of the biggest are bunker optimisation and empty container forecasting. They are multi-billion dollar cost areas with the kind of complexity the team thrives on.

Commercially, the ground is just as fertile, from automated and intelligent pricing and quoting systems to determining the most desirable cargo mix for vessel utilisation and customer profitability analysis.

New business models are a third area of the team’s focus, reflecting the likelihood that Maersk Line’s globe-spanning operational and commercial data – combined with the ability to analyse it – will one day be a valuable product itself.

“We don’t know where analytics will take us, but we can see the next few phases where this team will create a lot of value for the business,” says Jakob Stausholm.

“It’s early though, we’re learning to walk right now. We have to keep building our competencies, looking at the kind of tools we need and to define how we want to work. That includes giving the analysts room to play and explore their own ideas. It’s going to be exciting to see how it develops and where it takes us.”