FEATURES

Behind the DryDoQ Insights success story

Michael Hindmarsh, Incubator Lead in the UK, tells a tale of teamwork: the origin story of DryDoQ Insights. It’s a new tool to predict the corrosion and fouling status of ships to help with repair decisions for upcoming dry dockings.

Michael explains the idea for DryDoQ Insights first came out of a desire to find more ways to put our existing data to work for our customers in the marine industry.

A treasure trove of data

Our Marine business has already used collected coating performance and surface roughness data to develop Intertrac Vision, a digital tool that helps ship owners understand the impact of coating selection and surface preparation on a ship’s hull performance. But Michael and his colleague Richard Ramsden, Data Science Lead – Innovation, really felt there was more we could do with other collected data.  

Michael Hindmarsh, Incubator Lead in the UK and Richard Ramsden, Data Science Lead – Innovation
Michael Hindmarsh, Incubator Lead in the UK and Richard Ramsden, Data Science Lead – Innovation

“We decided to run a workshop led by the Marine business. The focus was to see if we could develop a value proposition for marine customers using corrosion insights from the large dataset of vessel inspections that our Marine business had collected over a 40-year period,” Michael says.

“We knew that when we collected fouling control performance data, that we also recorded corrosion data,” Michael continues. “We decided to see if we could generate corrosion prediction algorithms, and further turn them into a tool for giving customers a view of the status of their ship’s underwater hull. This knowledge would allow them to make decisions on needed repairs ahead of a forthcoming dry dock.”

Solving a problem

There was an opportunity here because the dry dock process, wherein a ship is removed from the water for maintenance or repair, has not changed much in many years. Arrangements are made months in advance, and often rely solely on experience and guess work to judge the extent of repairs required.

Michael adds: “Only when the vessel actually comes into the dry dock and the water is removed, does the full extent of the repair work become apparent. We felt we could improve that situation with our data.”

Combining expertise

Michael says that to fully understand how to proceed, they needed to explore the idea from a variety of angles: “This included thinking differently about the problem, and making sure that we reached out to experts and potential users in the marine industry.”

They held a facilitated workshop with a group of market, technical and sales experts, as well as invited external experts in data analysis, machine learning, predictive modeling, data handling and software design, and our knowledgeable customers – ship owners themselves.

“We believed the insights we gained could help us develop a digital product or offering for the market in a short timeframe,” Michael says. “It turned out to be a very worthwhile exercise. I admit, I was surprised at the level of engagement. We identified the roles and contributions of each participant, and moved quite quickly through the process. But what was most exciting was how quickly we embraced a lean startup type approach. Within weeks, we were interviewing marine customers to find out if the problem we had identified really existed. We literally got out of the building, out of the internal day-to-day activities and met the marine market potential end users.”

Flexibility enhanced the solution

It was not all smooth sailing, however. The main challenge, Michael points out, was that we wanted to develop a solution for an existing problem that was frustrating but not intolerable for our customers. Would they decide to invest in the solution, or not?

Michael says the team found a way to use this “challenge” to enhance the desirability of our solution. “We analyzed the problem interviews with potential customers, and noted other functions the customers were asking for. This really showed the value of getting out of the office early in the process and meeting with customers or potential end-users. We were able to ‘pivot’ the project solution to include those functions, like fouling challenge status to complement the corrosion status of the vessels.” 

“The way we handled the customer interviews demonstrated the extent of the skills and understanding that we already have within AkzoNobel, and how we add value by working with others,” Michael says. “For example, we had some of the external partners of the initial workshop help us with the customer interviews.”

He continues: “Another important aspect of this project was how we identified the gaps in our data handling skills – but have been able to overcome them. By showing openness in our developing capabilities, we have successfully engaged with external data science experts such as the National Innovation Centre for Data in Newcastle, UK, and through these relationships, we have been able to further develop our internal data handling capabilities and our understanding of machine learning. We are very proud that this close collaboration has meant that we are producing the final algorithms.”

Working together is leading

The learnings from this experience seem endless for Michael, but what stands out is what can be gained from working together on projects like this one.

“I’ve learned that we shouldn’t be afraid to take a lead in the industry. I’d encourage like-minded organizations to explore the benefits of working together – what we’re working on separately could very well have synergies which open doors for us both,” says Michael.