Search

This paper proposes a novel model-based approach to technology planning and roadmapping, consisting of two complementary steps: technology forecasting and game-theoretic planning. The inherent uncertainty of target technology performances, timelines and risks impact the roadmapping process. Reducing this uncertainty is a major challenge and allows elaborating different options and scenarios. A formal methodology is proposed for quantitative forecasting in a multi-dimensional space (different performance metrics and time) based on past technology development trend information. The method adopts concepts and approaches from econometrics and is formulated as a convex optimization problem with different constraints on the frontier's shape. It provides useful product line assessment benchmarks and helps to set reasonable goals for future technology developments. Game-theoretic planning allows addressing the strategic decisions to take, considering the technology landscape, markets, and competition. The strategic decisions affect in turn other companies as well, which is the basis for the application of game theory, in the form of best-response functions to determine the subsequent reactions and movements of rivals in a technological landscape. The result is a simulation of a sequential game in technology space, allowing evaluating possible technological development pathways and determining optimal models on the Pareto frontiers, potential targets for technology roadmapping. [ABSTRACT FROM AUTHOR]

• We propose a novel model-based approach to technology planning and roadmapping. • We consider both technology forecasting and game-theoretic planning. • We formulate a formal methodology using concepts from econometrics and game theory. • We demonstrate the approach on an automotive use case. • Our work provides a quantitative basis to support human decision making in roadmapping activities. This paper proposes a novel model-based approach to technology planning and roadmapping, consisting of two complementary steps: technology forecasting and game-theoretic planning. The inherent uncertainty of target technology performances, timelines and risks impact the roadmapping process. Reducing this uncertainty is a major challenge and allows elaborating different options and scenarios. A formal methodology is proposed for quantitative forecasting in a multi-dimensional space (different performance metrics and time) based on past technology development trend information. The method adopts concepts and approaches from econometrics and is formulated as a convex optimization problem with different constraints on the frontier's shape. It provides useful product line assessment benchmarks and helps to set reasonable goals for future technology developments. Game-theoretic planning allows addressing the strategic decisions to take, considering the technology landscape, markets, and competition. The strategic decisions affect in turn other companies as well, which is the basis for the application of game theory, in the form of best-response functions to determine the subsequent reactions and movements of rivals in a technological landscape. The result is a simulation of a sequential game in technology space, allowing evaluating possible technological development pathways and determining optimal models on the Pareto frontiers, potential targets for technology roadmapping. [ABSTRACT FROM AUTHOR]

With the increasing availability of data for Prognostics and Health Management (PHM), Deep Learning (DL) techniques are now the subject of considerable attention for this application, often achieving more accurate Remaining Useful Life (RUL) predictions. However, one of the major challenges for DL techniques resides in the difficulty of obtaining large amounts of labelled data on industrial systems. To overcome this lack of labelled data, an emerging learning technique is considered in our work: Self-Supervised Learning, a sub-category of unsupervised learning approaches. This paper aims to investigate whether pre-training DL models in a self-supervised way on unlabelled sensors data can be useful for RUL estimation with only Few-Shots Learning, i.e. with scarce labelled data. In this research, a fatigue damage prognostics problem is addressed, through the estimation of the RUL of aluminium alloy panels (typical of aerospace structures) subject to fatigue cracks from strain gauge data. Synthetic datasets composed of strain data are used allowing to extensively investigate the influence of the dataset size on the predictive performance. Results show that the self-supervised pre-trained models are able to significantly outperform the non-pre-trained models in downstream RUL prediction task, and with less computational expense, showing promising results in prognostic tasks when only limited labelled data is available. • Self-supervised learning is investigated to address data scarcity in prognostics. • Fatigue crack propagation prognostics in structures based on strain data. • Synthetic dataset composed of multivariate run-to-failure strain time series. • Self-supervised pre-trained models improve RUL estimation performance. [Display omitted] [ABSTRACT FROM AUTHOR]