The promise of data-driven asset management is optimal, cost-efficient area management. In other words, an ideal balance between risk, performance, and costs. We achieve this by intelligently collecting data. From there, we intelligently distill the information to ultimately predict which measures are needed and when. It may sound like an ambitious goal, but reducing surprises (failures) is certainly within reach.
Data-driven asset management isn't just about understanding maintenance needs, but also about understanding the organization. Data from the supply chain is necessary to gain insight into the status of work orders, costs, prices, and schedules, among other things. Data to inform asset-related work processes is essential to transition from data-driven maintenance to data-driven asset management.
Collecting data
Much has already been written about how to collect data. As a starting point, the websites IM Safe and their Wiki are recommended. This is a European initiative aimed at contributing to the standardization of infrastructure monitoring for optimal maintenance and safety. The CROW Monitoring Handbook, designed to facilitate the selection and deployment of sensors and recording equipment, is also of interest. Move beyond the trial phase and make data-driven asset management part of the regular management process.
In recent years, significant efforts have been made to develop data collection methods. These include detailed and accurate monitoring of energy consumption, vibrations, pressure, temperature, and image quality. Significant strides have also been made to simplify the collection of data and information from the supply chain, particularly from within the organization and from external stakeholders. The major challenge lies in developing methods to convert data into information about the asset's condition. This information can then be used in (future) predictive models, thus contributing to data-driven asset management.
Anchoring in the organization
The time is ripe for implementing data-driven asset management. However, we shouldn't pretend it can predict everything right now and instantly implement the right strategy for an entire area. Instead, a process can be established that facilitates a growth path for data-driven asset management. A process with room for further innovation, experimentation, evaluation, and learning. It should be so firmly embedded in the organization that the process becomes part of regular management processes. Management processes are often based on the PDCA cycle. The PDCA cycle offers an excellent framework for development, and thus also for data-driven asset management. Developing predictive models, for example, can be part of the regular management process. So, not a separate program, project, or pilot, but a growth process for data-driven asset management.
It's often said that if you don't know the failure behavior of a component, you don't know what data to collect. But the reverse is also true: if you don't collect data, it won't provide any insight into the failure behavior. Both arguments are valid. It's strongly reminiscent of the chicken and the egg story. Our proposal: choose the chicken and the egg. Collect the data we already know is necessary, any data that can be obtained relatively easily as by-catch, and data on usage and circumstances. This will create a comprehensive, usable dataset over time.
Interpreting field data and converting it into usable information for management and maintenance is often part of the management process. Only the methods need to be adapted or expanded to implement data-driven asset management. For example, by plotting malfunctions, measurement, and inspection results on a dataset of usage and conditions, a predictive model can eventually be developed. This type of data must be analyzed within a management process anyway, so it can easily be incorporated. Embedding it in the management process ensures continuous improvement of data, information, and predictive models, and thus also of data-driven asset management.
What does it yield?
What will the implementation and development of data-driven asset management cost? It's an easy question, but difficult to answer. However, the (social) costs resulting from current working methods, potentially with insufficient maintenance and likely significant inefficiencies, are also unclear. Or should we focus on the bottom line? Significant savings by preventing capital destruction due to inadequate maintenance, reducing operational disruptions, the intrinsic value of the knowledge gained, and performance tailored to needs.
Want to know more?
Is your organization working on data-driven asset management or would you like to discuss this with one of our professionals? Feel free to contact us. We're happy to put our knowledge into practice!
