Driving data quality and openness

Vessel Performance Solutions has developed a performance analysis tool that is already attracting attention from ship owner clients far beyond its native Denmark, where the government is investing in the company under the INNOSplus project to improve shipping information systems. According to founder and owner Jakob Buus Petersen, the Danish stakeholders view an open data exchange standard as a crucial development

“If half your fleet are on an ABB system and half on Siemens, how do you analyse their performance through one system,” explains Buus Petersen. “Ship owners don’t want to be stuck with just one supplier. The requirement for an open standard comes from ship owners, and the Danish government see a big competitive advantage for their ship owners in being the first to develop a truly open exchange format.”

The development of the open standard will be part of the next phase of Vessel Performance Solutions’ work. In the meantime, another thorny issue is under attack. The proliferation of sensor data onboard modern vessels comes with an important caveat – it is often the quality, not just the quantity of information that is important.

This importance is long documented but little appreciated; Buus Petersen cites a study from 1998 (Redman) that estimates that poor data quality can cost a typical organisation between 8-12% in revenues, and may account for up to 60% of a service company’s expenses. But what is poor data, and what is good enough? There are few established answers, but VPS proposes a hierarchy ranging from those that are easy to achieve and have a small impact (such as timeliness for some applications) and those that are more difficult to achieve and have a bigger impact, chief among them consistency and accuracy.

What is needed, and what VPS is honing under the INNOSplus project, is a validation engine. This consists of many levels of rules applied to data output from a vessel’s (or an entire fleet’s) systems, ranging from a simple check for completeness to a more complex set of accuracy tests. Rules define parameters for data and trigger a report when they are met. By analysing the statistics on rules triggered, data quality can be quantified and problems addressed.

The project draws on the experiences of Danish operators J Lauritzen and Torm, who have both operated performance management systems for some time and are now investigating ‘data warehouse’ solutions to store, analyse and learn more from their operational experiences. And Buus Petersen reports that VPS is already working with several other major ship operators. As more ship operators address how to make the most of the data they generate, these questions of format and accuracy are set to become even more critical.