ABSTRACT
Predictive Maintenance (PdM) solutions can predict and consequently minimize downtimes. Concerning high implementation cost of PdM solutions for manufacturing systems as well as for sold systems operating in the field, e.g. Building Energy Systems in residential heating (RH), it is unclear under which conditions potentials outweigh investment. In a previous scientific publication, the authors proposed a methodology to systematically assess potentials and benefits of to-be-developed PdM solutions in manufacturing processes. The utilization of the methodology for RH systems is promising, but requires a suitable adaptation to its individual context. Apart from quantifying the benefit compared to investment for PdM solutions, the context of RH is fundamentally different from manufacturing. As a major difference, the systems are not located in-house, but in the field at the customers’ site. This leads to increased challenges in terms of lacking customer and system feedback and data availability. Therefore, this paper addresses the optimization, transfer, and adaptation of the previously developed methodology to RH systems. Additionally, this paper includes baseline information regarding maintenance and RH and an exemplary application of the optimized and adapted methodology that supports the reliability department of BOSCH Thermotechnology for quantifying the potentials of PdM solutions in the field of RH.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1. The MLC is a template for structuring a complex project. The MLC helps to describe and specify steps that take place in a ML project, e.g. what data the product is based on, what goals are pursued, what kind of machine learning is used, how the predictions are used, and how the system can evolve and adapt (Marin 2019).