Skip to content
Pierre Dersin ny bok
View original picture , opens in new tab/window

New book on innovative method of predicting failure

Published: 30 May 2023

How can we predict when machinery and systems will fail? In a new book, Adjunct Professor Pierre Dersin wants to build bridges between engineering disciplines to make better models of products remaining useful life. He hopes to further develop and apply the ideas in the book in collaborations within Luleå University of Technology, AI Factory and CIAM.

Accurate predictions of how long machinery and systems will work according to requirements is important both to avoid failures and to keep maintenance costs reasonable. Pierre Dersin has many years of experience from working in the industry, mainly in railways, and found himself in the interface of two disciplines both important for products and systems aging and failure – Reliability engineering and Prognostics and Health Management.

“People within the discipline reliability engineering and the discipline prognostics and health management did not talk much, not only in my company but in general. I became interested in uniting the areas and combining methods. This was the base of some publications, and I was then persuaded to write a book,” says Pierre Dersin, Adjunct Professor of Operation and Maintenance at Luleå University of Technology.

Calculating remaining useful life

The book is about an important indicator related to failure prediction – remaining useful life (RUL), meaning the time left that a product or system will perform normally. The easy way of calculating remaining useful life would be for something that ages linearly until failure. However, this is not usually the case in the complex reality. Pierre Dersin presents an innovative method for deriving a nonlinear time transformation which enables the average RUL to become a linear function of time. Then useful reliability engineering properties can be exploited to describe the dynamics of RUL, how it evolves with time and to quantify the uncertainty in predicting the indicator.

The book also includes statistical inference methods, and maintenance optimization.

“If you know how fast something is aging, you can propose the best maintenance policy, taking into account remaining useful life and the risk of failure you are willing to take,” says Pierre Dersin.

Welcomes discussions and new applications

The book is aimed at researchers, students, and maintenance practitioners in the industry. Pierre Dersin hopes that partners at AI Factory and CIAM at Luleå University of Technology will find applications within their fields.

“I look forward to discussions with colleagues with slightly different backgrounds, cross fertilization is always fruitful. I also hope this will be useful for the industry, there are many potential applications in addition to those described in the book”, says Pierre Dersin.

Pierre Dersin

Pierre Dersin, Adjunct Professor

Organisation: Operation and Maintenance, Operation, Maintenance and Acoustics, Department of Civil, Environmental and Natural Resources Engineering