果冻传媒

Effective asset management is crucial for reducing costs related to the acquisition and maintenance of capital assets. Companies must smartly plan and execute maintenance activities to ensure the reliability and longevity of capital assets. Informed decisions about the type and timing of maintenance have a significant impact on operational efficiency and cost savings. In his PhD study, Ragnar Eggertsson conducted research on maintenance optimization under incomplete information and sustainable technology selection.

Maintenance optimization under incomplete information

Incomplete information arises when a system's state is monitored incompletely, for instance, by a sensor or if inspections are performed infrequently. Eggertsson's work considers maintenance optimization under incomplete information in three different settings:

1. Maintenance planning for Heating, Ventilation, and Air Conditioning (HVAC) units in trains of Dutch Railways (NS). The research incorporates environmental dependency of a train's HVAC condition information in inspection and maintenance optimization. This can lead to more than 10% cost savings.
2. Another example is maintenance optimization for a multi-component system with a single sensor. This example originates from a case study of a printer. The sensor provides partial condition information, which helps make maintenance decisions. The study concludes that an optimal policy with up to three decision areas can be developed under specific conditions.
3. Inspection, maintenance, and replacement optimization for sewer pipes. In collaboration with Rolsch Assetmanagement he conducted a case study based on data from the Dutch city of Breda. They showed that an optimized condition-based maintenance policy greatly improves policies currently used in practice. These perform at least 34% worse than the optimized policy.

Sustainable technology selection

Power producers face the challenge of navigating the energy transition by changing their energy-production portfolio. Currently, many power producers own carbon-emitting power plants, but to reach climate goals, these need to be phased out in favor of clean energy sources. Eggertsson's research studies how a power producer operating in a competitive market, switches between technologies based on exogenous price processes such as electricity, coal, natural gas, and carbon prices. With a case study based on real-world data, he highlights the need for smart implementation of policy measures such as carbon pricing.

Key takeaways

• Data-driven, context-specific maintenance policies outperform generic maintenance policies where they can be applied.
• It is important to consider information incompleteness when optimizing maintenance for a capital good, as inspections and maintenance can otherwise be incorrectly timed, leading to performing these actions too often or, worse, unexpected failures.
• When pricing carbon, policymakers should consider long-term power plant switching behavior; for a transition to renewable energy, a switch to a cost-competitive intermediate technology might hinder the transition to renewable in the long run.

Eggertsson's research highlights the importance of data-driven, context-specific maintenance and technology selection decisions in asset management, providing valuable insights for professionals and policymakers alike.

Ragnar Eggertsson defended his thesis on November 22. Title of PhD thesis: "Advances in Asset Management: Maintenance Optimization under Incomplete Information and Sustainable Technology Selection." Supervisors: Rob Basten and Geert-Jan van Houtum.