Êtes-vous un étudiant de l'EPFL à la recherche d'un projet de semestre?
Travaillez avec nous sur des projets en science des données et en visualisation, et déployez votre projet sous forme d'application sur Graph Search.
Motivation Over the past 30 years, within the field of infrastructure management, as infrastructure systems have grown more complex, the level of managerial and financial oversight has increased, and the computation power has become less expensive and more readily available, infrastructure managers have increasingly turned to computerized infrastructure management systems to aid in the management of the built infrastructure. An infrastructure management system acts as an infrastructure manager's 1) inspection database detailing the current state of the built infrastructure, 2) infrastructure performance simulation platform modeling the potential future infrastructure performance, 3) infrastructure deterioration and provided service estimator evaluating the equivalent financial loss associated with infrastructure deterioration and decreased provided levels of service, and 4) infrastructure maintenance action development engine formulating and scheduling technically and financially optimal infrastructure maintenance solutions. Historically, the performance simulation module of infrastructure management systems has focused primarily upon modeling gradual infrastructure deterioration processes, such as corrosion, and the infrastructure provided service estimator has employed nominal values for quantifying the public's evaluation of the provided infrastructure performance. These approaches have caused infrastructure managers to focus their attention and funding towards combating gradual infrastructure deterioration while giving priority maintenance status to infrastructure objects that place the largest nominally evaluated total performance impact on society. While this approach has helped infrastructure managers to more efficiently manage the built infrastructure, such an approach is only efficient if it is unfailingly implemented over multiple decades. Unfortunately, the current limited infrastructure management system scope has exposed infrastructure managers to two potentially disruptive forces – potential unforeseen natural hazard induced technical failures and potential political and/or financial funding support failures due to incongruent evaluation of the provided level of performance between the infrastructure manager's nominal evaluation measures and the experiencing society. While these two disruptive forces originate in two very different elements of an infrastructure management system, they both can induce the same result – undermining of the intended technically and financially optimal infrastructure management solution. Objective and originality To work towards rectifying both of these limitations, the current work has developed methodologies for both quantifying the long-term infrastructure natural hazard risk exposure and estimating an individual's experience-based evaluation of the provided level of infrastructure performance. The methodology for assessing natural hazard induced technical infrastructure failures has focused around developing an infrastructure component potential failure assessment procedure which employs data within existing infrastructure and transportation management systems and currently under development natural hazard identification maps. This failure assessment procedure is designed to identify the locations and estimate the associated consequences of potential natural hazard induced failures. This failure probability and failure consequence information is then employed to quantify an annualized and a multi-year infrastructure link risk exposure. The methodology for estimating an individual's experience-based evaluation of the provided level of infrastructure performance has been developed by employing research findings from the fields of psychology and behavioral economics to develop the affective assessment approach, a variance-based evaluation tool. This evaluation tool employs an individual's range and quality of experienced infrastructure performance in time to predict the individual's current induced affect and future sensitivity to the provided levels of infrastructure performance. As this is a constructed evaluation approach, employing the affective assessment approach directly in the third infrastructure management module, evaluating equivalent financial loss associated with decreased provided levels of service, can be computationally prohibitive as each potential sequence would need to be modeled. It is, therefore, proposed to use the affective assessment approach as a reality check against which the developed technically and financially optimal management solution can be assessed and solutions which provide too low or inconsistent levels of infrastructure performance can be identified, reevaluated or discarded. Results and benefits With this additional technical information and perspectives, infrastructure managers are able to actively consider an object's natural hazard risk exposure in modeling the infrastructure deterioration and in developing optimal maintenance actions. Furthermore, infrastructure managers are able to determine the annual funding that should be invested and made available to respond to current and future natural hazard induced object failures. By, calibrating and implanting the affective assessment approach, infrastructure managers can also study how proposed technically optimal solutions may be socially received and can select solutions which best maintain social support throughout the duration of the maintenance solution. Within the field of infrastructure management, the past 30 years has been invested in developing standardized and computerized infrastructure management systems. With these systems implemented, infrastructure managers are starting to observe their strengths but also their limitations. This work has focused on formulating quantitative methods to develop solutions for two of these limitations. It is hoped that these methods might be further developed, calibrated and implemented to improve existing infrastructure management systems and the infrastructure and society they manage.
Anton Schleiss, Daniela Rodrigues, José Pedro Gamito de Saldanha Calado Matos