I recently attended a conference hosted by the Rail Alliance on the subject of Whole Life Costing (WLC) in the rail industry. The event seemed primarily focused at HS2 and what lessons could be learned from across the industry for their benefit, but it raised interesting questions for other infrastructure organisations too.
Just looking around the room, it was encouraging to see so much interest and enthusiasm from across the supply chain. The day, however, confirmed for me how data quality plays an absolutely critical role in WLC analysis. Although many advanced techniques can be applied to WLC analysis to factor in more, and increasingly complex, variables (at levels of mathematics far beyond my own capabilities) I think it boils down to a far more simple equation:
Garbage in = Garbage out.
I know this may seem a cliché, but it’s important for a good reason. The presenters from Whole Life Consulting (both very knowledgeable, accomplished speakers) recognised a potential for a 30% error in the outcome of WLC analysis vs reality. 30% is huge! Although projected costs for HS2 are still variable, if the construction cost is £30bn and that represents, say, 20% of Whole Life Costs, then the errors in WLC based on poor data could be £45bn.
Historically, the rail industry has not been great at managing its data, though progress has been made in recent years with the realisation that data is a key enabler to improving asset management. For the purposes of WLC analysis it would certainly be worth acquiring confidence grades in data quality before going too far, in order to minimise the range of potential error. If the quality of data is better understood, then it should be possible to make assumptions about the impact on WLC analysis and ensure that a suitable tolerance range is quoted, for example WLC is £x +/-y%
In order for WLC to have a realistic degree of accuracy and reliability, it will require reliable data both from the supply chain about their products and from established rail organisations on their historic asset data. I am certain that WLC is the way forward to understand and demonstrate the benefits and value of asset investment, but I believe there is a need to set realistic expectations which factor in data quality issues in the analysis.
What high value impacts of poor data are you aware of?