Skip to main content

AkzoNobel - Case Study

AkzoNobel - Case Study

How a global company learned to harness its data – and create products for the future.

Building predictive models with global paint experts

AkzoNobel is a global expert in paint and coatings. Its products:

  • decorate homes and businesses
  • protect pipelines and turbines
  • coat aircraft, automotive vehicles and even marine vessels

Working with us, AkzoNobel’s representatives built predictive models. These provide customers with vital business insights, as well as a quality product.

They also learned some of the skills that would allow them to turn more of their ideas into reality in the future.

This is a real transformation for AkzoNobel; to develop from a company that sells paint, to a company that provides APIs and data.

Richie Ramsden, Data Insight Team, AkzoNobel

The goal

AkzoNobel has a history going back hundreds of years. But even the most established companies have to keep their eye on what’s coming around the corner.

In setting up its Innovation Incubator, the company pledged to back promising ideas. This was particularly true in areas where uncertainty and risk were high.

They aimed to develop innovations that would make up a large proportion of turnover by 2025.

The company approached us with two goals in mind, to:

  • help it use its data to create models that could benefit customers
  • learn some of the skills that would enable it do more of this work in-house

The results

With our help, AkzoNobel representatives undertook a six-month project at Newcastle Helix. In that time, they:

  • cleaned and analysed data from different locations, and in different formats
  • developed a minimum viable product that could offer real value to customers
  • learned a variety of skills in predictive modelling and machine learning
  • encouraged other AkzoNobel staff to consider how they could develop their own ideas

From “descriptive” to “predictive” data

AkzoNobel is no stranger to collecting data. It collects business intelligence from many different markets and sectors. These go into descriptive analytics dashboards. They help teams track sales in volume and location.

It had worked with consultants on projects that pooled data into “predictive” models. This would help it the company make smarter decisions. But that arrangement meant that the company was reliant on external skill. They might have relied on products that could be sold, changed or shutdown.

“We had the ideas, but outsourced to others to do that work”, says Richie Ramsden, from AkzoNobel’s Data Insights Team. “We decided a project with NICD that could help us upskill – or at least understand some of the skills – was vital.”

They even had a project in mind.

Richie Ramsden

Sorry, you need JavaScript to view this video

Corrosion in motion

When you’re operating a fleet of marine vessels travelling the world, you’ve got to take care of your hulls. AkzoNobel sells anti-corrosive coatings. They help to protect hulls from everything from salt water damage to barnacles.

These coatings can last as long as 25 to 30 years. But ships still need to check back into dry dock every five years or so to assess how the vessel is performing. Vessel operators want to keep time in dry dock brief, so they can save time, money and fuel.

So AkzoNobel needed a solution to allow customers to predict how their hulls were holding up. One that didn't involve having to wait for a docking or dive inspection.

Needing a collaborator

The company had worked on a similar project in the past. It had extensive insight into the factors that caused corrosion. It even had a range of datasets it could draw on. These included the positioning of 80,000 vessels, taken every 15 minutes for the past nine years.

What it needed was a collaborator that could enable it to understand and prepare the data it needed.

“We have data scientists in house. But we realised it was a much larger project than building predictive models”, says Richie. “It was about data engineering, bringing data from multiple places, understanding the cloud aspect of the project and scaling the computing.”

AkzoNobel supports Newcastle University’s EPSRC Centre for Doctoral Training in Cloud Computing for Big Data. It trains future leaders in data analytics. So they knew the benefits of working with people that had a strong balance of skills.

Richie said: “70 to 80% of your time on projects like these is on cleaning and quality controlling the data. That’s why the skills at NICD are so crucial.”

A "critical friend"

The company met with NICD in February, and started the project at Newcastle Helix in mid-April. NICD Deputy Director Barry Hodgson says NICD’s role in the project was as a “critical friend”. This was rather than a conventional research partner or consultant.

He added: “We’re not the consultant for hire. It’s a co-creation relationship. We’re the friend that works with you and says: ‘That’s good, but have you thought about it this way?’. It’s a collaboration, but ultimately it’s led by the company, and what it needs.”

Early in the project, Richie held a team meeting at Newcastle Helix. He brought members of the Incubator in to brief them. He invited them to ask questions of the NICD team.

Richie added: “That was really crucial for getting buy-in. People come in and see the skills that are available..”

We’re not the consultant for hire. It’s a co-creation relationship. We’re the friend that works with you ...

NICD Deputy Director Barry Hodgson

DRYDOQ emerges

The idea was to train a core team as part of the project. Richie was keen to bring as many people on-site as possible to experience what was happening.

“We had people based at Helix who we wanted to upskill. But we brought several people in to sit with the team and absorb some of the language and knowledge”, says Richie. “Some people were trying to change the way their departments used and collected data. So they’re exposed to this sort of work, and then go away and start thinking about what else might be possible.”

Meanwhile, the team learned more about the data that was available and how they might use it. They looked at formatting of data, and how they could share it securely and seamlessly.

“When we started, we thought we might take what was already created, and scale it up. What actually happened was that we almost completely re-built it from the ground up. We created something even better.”

Two models for solution

By September, they had created two models.

  1. Enable customers to gauge the corrosion status of a hull, presented in a red, amber or green status.
  2. Present level of corrosion as a percentage for customers to work out paint required in dry dock.

These models form the basis of DryDoQ Insights. It's a predictive tool that helps AkzoNobel customers. It enables them to make better decisions about how to maintain and protect their fleets. The company began testing the minimum viable product at the start of the year.

“We’re providing Vessel Operators with insight into their vessels”, says Richie.” We’re also talking to adjacent customers, from dry docks to ports and people who charter vessels. Instead of a transactional business model, we offer opportunity to make data-driven decisions.

“This is where a lot of businesses are going now. It’s not just about selling products. It’s also about selling what the product does, and helping customers understand it better.”

We’ve learned about machine learning and data analytics. We know what goes into running this kind of project again.

Richie Ramsden, Data Insight Team, AkzoNobel

Future ideas are born

The collaboration with NICD didn’t just lead to the creation of a new project. It provided AkzoNobel with more understanding of the skills involved in creating products.

Richie says: “We’ve learned about machine learning and data analytics. We know what goes into running this kind of project again. We also know what abilities we need. We can put together a specification to recruit for people with these kinds of skills.”

In January 2019, AkzoNobel launched Paint the Future. It's a start-up challenge for institutions, teams and experts to work with the company. They develop ideas, improve processes and come up with inspiring solutions.

Through work with NICD, Richie says the Data Insights Team has more understanding. They know how to secure and share data, and make the most of these link-ups. The excitement of the NICD project has also filtered down to other AkzoNobel staff. Other departments are looking at similar projects to deliver efficiencies and new products.

More projects

“Saying we’re working with the National Innovation Centre for Data is a big thing. And that’s borne out by the way that people are exploring other projects as well. This has been successful and we’ve got something out of it. Others in AkzoNobel are seeing those positives and talking about running other projects.

“If you talk to anyone in research, sales and marketing, everyone’s talking about how they can use data better. They’re looking at us, and asking what skills they need to do something like this.

“With NICD’s expertise, we’ve used a wealth of data to build digital products for the future. That feels really exciting to me.”