From empirical and subjective O&M to statistical and objective asset management
This lecture will focus on the Danish Road Directorate’s (DRD) ambition for future asset management across all of DRD’s asset classes and describe the specific opportunities and challenges this entails for the future asset management of DRD’s structures.
DRD is responsible for over 3800km of national road, which consists of a multitude of different assets such as roads, bridges, tunnels, drainage, noise barriers, utilities, signals, bus shelters and much more. These assets have until now been managed under different assumptions and in separate processes and IT-solutions, obfuscating the overview across assets and limiting the opportunities to optimise operation and management (O&M) budgets. DRD’s needs for O&M funding are reported yearly to the government and are then part of the coming finance act (finance bill) debated and approved by the Danish Parliament. Depending on the state of the country and current affairs, the budgets coined out to DRD’s O&M may vary from the re-ported needs. DRD’s responsibility is therefore to ensure that decision makers are informed about the loss incurred by not meeting the reported needs and ultimately to ensure that the provided budget is leveraged to provide the most beneficial effect in the entire road network, i.e. best value for money given all known conditions.
As assets have different performance requirements and degradation curves, the cross-asset optimisation has until now in parts been handheld and based on subjective evaluation. The development of data compu-tation has however opened for a possible major improvement of this approach, so that more automated, data-driven and objective reasoning can lead to improved maintenance strategies across the different asset classes. In order to do this, DRD is setting out and firming up guidelines to ensure that assets are treated under same constraints and to ensure that needs from different assets can be compared and aggregated to provide an optimum cross-asset strategy.
When available budgets differ from the needs, the future asset management system shall revert the analysis of needs and identify the preferred sub-optimal solution that ensures minimum monetary loss. The loss is calculated as the loss from maintenance not being performed at the right time leading to increased mainte-nance costs later; or the loss because maintenance is carried out too early resulting in the loss of the remaining lifespan of an asset.
In addition to the above, the development of data collection (e.g. drones, photogrammetry, damage detec-tion, structural health monitoring, etc…), data capacity and machine learning, have opened up for new op-portunities to analyse the remaining lifespan of assets and improve our understanding of optimum mainte-nance strategy, especially for structures. DRD’s current practice for assessing the condition of structures is an empirical model that has been well tested over the last 40 years and is grounded in qualified but subjec-tive engineering assessments. The focus going forward is to establish objective and statistical forecast to get better precision. This in turn, however, will require stricter measures on data quality, determining data that has statistical significance, and thereby defining the type of data collection DRD should carry out in the future.
The lecture will delve into further details on the above-mentioned considerations and provide examples on how DRD’s asset management of structures is set to change.