Over the past six months we have been working with the reliability and maintenance organization within a large oil and gas client to build out their Master Equipment List (MEL). Like many asset-intensive organizations, they have implemented an Enterprise Asset Management (EAM) tool to give them visibility and control over their capital equipment to optimize their maintenance strategies, reduce operating costs, and better manage their workforce and spare parts inventories. The challenge is that the software tool was implemented many years after the assets were constructed/procured. The data – the MEL – has to be built. Achieving the benefits of an EAM tool rely on a complete and detailed MEL populated with all critical pieces of information – type of equipment, manufacturer, model, serial number, vendor, criticality, spare parts listings.
The project sounded quite straightforward when we provided the bid. Find the missing data – major things like manufacturer, model, serial numbers, vendor information, purchase orders – for their offshore equipment and populate the Enterprise Asset Management (EAM) tool. At Sullexis, we are data experts. We know how to merge it, scrub it, format it, load it. We are accustomed to pulling data from a complex array of sources and defining a common model. We know how to develop extract, transform and load (ETL) routines. We know a lot about EAM tools, reliability and maintenance strategies, and the oil and gas industry. This seemed like a very simple request.
But it wasn’t. We have been working to find data on assets that were built/procured 10+ years ago. The POs are not clear, the vendor documentation is inconsistent and few of the people who worked on these original projects are still in the same roles. There were many companies involved in these large projects. The concept of “information management” didn’t really exist 10 years ago in any of these companies. Because the equipment is offshore, it is not feasible to actually walk down the information. We needed to find the original data sources.
And thus began our role as a private detective agency. We began a series of interviews to reconstruct the creation of these offshore platforms. Who did the design work? Who did the procurement? What artifacts were produced in those activities? Who did they transition their data and documents to? How was the equipment commissioned and who was responsible? Who were the vendors to build the skid packages? How was documentation stored in the shared drives and/or in document control? These interviews often begin “Where were you in 2005 when this platform was in the design phase?” and continue to “Do you know the location of any data or documentation related to this equipment? If so, where did you last see it?”
In this interview process, we have encountered the full spectrum of interviewees. The eager to help, but know nothing. The reticent types who respond to an essay question with a yes/no answer. The naysayers who just laugh and say it is an impossible task and we should give up now. The naysayers also like to spend most of the meeting time explaining all the reasons why we will fail and why every other team has failed in this endeavor, which actually makes them helpful as we frantically take notes of every previous mistake to ensure we don’t repeat them. And finally, after we relentlessly pursue meetings with every possible person who might be able to help us, we stumble across a few genuinely knowledgeable people with an actual memory who help us.
Through these interviews, we were able to identify a set of systems that contained a small piece of the overall puzzle we needed to assemble. The final data sets had to come from the documentation. And it was massive. Our team has had to filter through 100,000+ documents. Some of these documents are 1,000s of pages long and are typically low quality scanned PDFs that are not searchable and not well indexed. Some are even hand-written – and we’re not planning to give any gold stars to this team for penmanship. We had to manually review these documents to capture the relevant equipment data into spreadsheets to feed our data analysis team.
But our persistent reverse engineering of a set of major projects over a decade ago is paying off. We are succeeding. We have already identified over 40,000 sets of equipment data, and we have started the spare parts phase of the project. And – music to a consultant’s and a client’s ear – we are on track to meet our original budget and timeline.
If your organization needs help to define your MEL as part of an Enterprise Asset Management implementation, make sure your data team includes some private detectives.