Effective infrastructure asset management is crucial to optimise operations, reduce costs and achieve sustainable organisational growth. For an introduction to optimised infrastructure asset management, check out our insight on the benefits of mathematical modelling in infrastructure asset management.
In this article, we will outline how forecasting in infrastructure asset management enables predictive maintenance, more effective resource allocation, and improved data integration. We will also explain how forecasting can be applied when regular, high-quality information about asset health is not available.
Using historical data and predictive models, you can develop accurate forecasts of each asset’s condition, failure probability, and maintenance needs. This means you can proactively repair or replace assets and develop contingency plans to minimize disruption and risk to the wider infrastructure. This is especially important when asset failure can cause costly damage to further infrastructure – or even endanger lives.
Forecasting is also key to long-term operations. By maximising asset value and minimising risks through proactive forecast-driven planning, you can reduce the resources dedicated to dealing with emergency repairs. In addition, predictive asset models can help you to adjust strategies as new challenges emerge ensuring that they remain effective and aligned with future requirements and goals.
An example of the advantages of forecasting in infrastructure asset management are submarine data transmission cables. These cables are fitted with repeaters to boost signal strength. Should a repeater fail or a cable be damaged, the cables must be retrieved from the seafloor, repaired, and re-laid. But these cables can be kilometres undersea, so not only does their repair interrupt vital communication networks, but it also entails high-cost work in challenging environments. If you could forecast when cables were more likely to fail or be damaged, you could reduce disruption by making contingency plans for proactive maintenance, and even plan investment in new infrastructure.
Predictive maintenance to optimise asset management
Predictive maintenance is a data-driven solution to asset management. Traditional asset management is either reactive or preventative: assets are either fixed when something breaks, or inspected and maintained on a schedule. This can be costly and inefficient. By understanding how factors such as severe weather, climate, and frequency of use degrade an asset, predictive maintenance forecasts its condition and uses that forecast to determine the best time to intervene. For example, Network Rail use condition forecasts of their almost 200,000 rail earthworks to inform maintenance and renewal.
Accurate forecasting enables you to optimize your use of resources. By having foresight of how asset condition will change over time – and importantly, why – you can ensure that staff, equipment and materials are available when and where they are needed most. Furthermore, good forecasting can aid budgeting and help to make cases for future operation and investments. Best of all, assets can be used to the full extent of their lifetime when repair and inspection is constrained – like in nuclear reactors, where our forecasting model supports the safety case for continued reactor operation while avoiding the cost of being unnecessarily conservative and enduring excessive reactor downtime.
Future-proofing with data integration
Building a strong forecasting model demands the integration of diverse data sources. For example, maintenance records, sensor data and visual inspections, plus external factors like rainfall and climate change projections. While this creates complexity, it also drives collaboration and adaptability. The insights gained from asset condition forecasts can inform decision making processes and identify emerging trends so that strategies can be adjusted accordingly. This adaptability ensures that asset management remains effective and aligned with future requirements.
Cooperation between departments – operations, maintenance and finance, say – is essential to fully capitalise on insights gained from forecasting. Organisations need to align asset management strategies with the broader business objectives. Those that embrace data integration, department collaboration, and the advanced analytic capabilities of forecasting, will find that their asset management system is future proofed to changing technologies and business environments.
Forecasting with less data
In some cases, there is little available or useful data for infrastructure asset managers to use to develop models or forecasts. This can be for a variety of reasons:
- Assets are inaccessible
- Automatic monitoring of assets requires significant up-front investment in sensors and technology
- Data integration capabilities are lacking
- Acquiring regular data would be detrimental to operations
- Asset types do not support sensors
- Asset networks are too large to monitor regularly
These challenges require a different approach. We have worked on large networks of assets with Network Rail who have over 200,000 earthworks to manage. We have also worked with inaccessible assets, for example nuclear reactors with on projects with EDF, helping them to predict the future degradation of graphite bricks.
Incorporating a Bayesian approach that accepts uncertainties in asset data, and does not require regular, high-quality information about asset states means we can provide clients with challenging assets with insights to support their asset management strategies.
Smith Institute’s advanced mathematics knowledge can help our clients to harness the potential of their data when there is not enough available for other data analysis techniques.
We work with experts to incorporate prior beliefs about the assets’ behaviour to provide useful recommendations within a given risk appetite even when asset measurements are sparse.
How to get started with forecasting in asset management
Consider the following guidance when implementing forecasting as an asset management solution.
- Define clear objectives: How will forecasting support your asset management? What benefits do you want to achieve? For instance, optimising maintenance and downtime.
- Identify data sources and invest in data integration: Map the data you have available and consider what additional data might be needed. In parallel, invest in robust processes for data integration ensuring data is appropriately collected, cleaned, and stored. If there is insufficient data, consider how your approach will need to adapt.
- Plan for scalability: Ensure that the system you build is ready for future expansions and can incorporate evolving technologies.
- Prioritise change management and training: Implement a clear plan to ensure smooth adoption of your asset management system. Train employees to effectively use the forecasting results to make data driven decisions.
- Seek expert advice: Due to the technical nature of forecasting in infrastructure asset management, it be necessary to use out-of-house expertise.
By harnessing the power of available predictive analytics and historical data you can anticipate maintenance needs, optimise resource allocation, and minimise downtime. This means your assets operate at peak performance and deliver long-term value. With forecasting as a guiding tool, you can cut costs, turbocharge operational efficiency, and be ready for the future.
If you would like to understand more about making more effective infrastructure asset management decisions, get in touch here.