Search

Your search keyword '"Adey, Bryan T."' showing total 319 results
Start Over Author "Adey, Bryan T."
319 results on '"Adey, Bryan T."'

4. Implicit supervision for fault detection and segmentation of emerging fault types with Deep Variational Autoencoders

Author

Chao, Manuel Arias, Adey, Bryan T., and Fink, Olga

Subjects
Computer Science - Machine Learning, Statistics - Machine Learning
Abstract

Data-driven fault diagnostics of safety-critical systems often faces the challenge of a complete lack of labeled data associated with faulty system conditions (i.e., fault types) at training time. Since an unknown number and nature of fault types can arise during deployment, data-driven fault diagnostics in this scenario is an open-set learning problem. Most of the algorithms for open-set diagnostics are one-class classification and unsupervised algorithms that do not leverage all the available labeled and unlabeled data in the learning algorithm. As a result, their fault detection and segmentation performance (i.e., identifying and separating faults of different types) are sub-optimal. With this work, we propose training a variational autoencoder (VAE) with labeled and unlabeled samples while inducing implicit supervision on the latent representation of the healthy conditions. This, together with a modified sampling process of VAE, creates a compact and informative latent representation that allows good detection and segmentation of unseen fault types using existing one-class and clustering algorithms. We refer to the proposed methodology as "knowledge induced variational autoencoder with adaptive sampling" (KIL-AdaVAE). The fault detection and segmentation capabilities of the proposed methodology are demonstrated in a new simulated case study using the Advanced Geared Turbofan 30000 (AGTF30) dynamical model under real flight conditions. In an extensive comparison, we demonstrate that the proposed method outperforms other learning strategies (supervised learning, supervised learning with embedding and semi-supervised learning) and deep learning algorithms, yielding significant performance improvements on fault detection and fault segmentation., Comment: 22 pages, 9 Figures

Published
2019

7. A preliminary investigation of the potential benefits of using the ASTRA Bridge for short-span bridge deck refurbishment projects in Switzerland.

Author

Zumstein, Marco, Chen, Qian, Adey, Bryan T., and Hall, Daniel M.

Subjects
COST control, BRIDGE floors, BRIDGE design & construction, CONSTRUCTION costs, QUANTITATIVE research
Abstract

How bridge refurbishment projects are performed requires a trade-off between the speed and cost of the project and the amount of traffic disturbances during the project. A possible way to help reach a better balance between these two extremes is the ASTRA Bridge developed in Switzerland. The ASTRA Bridge is a 236-meter long steel ramp system on wheels, which is placed on top of the bridge deck undergoing refurbishment to enable vehicles to continue to pass over the bridge while construction work progresses underneath. This study illustrates new refurbishment processes by using the ASTRA Bridge and presents the first quantitative analysis of the effects of using the ASTRA Bridge on the time, costs and traffic disturbances associated with bridge refurbishment. The bridge investigated is a short-span (50 m long) highway bridge requiring refurbishment of its superstructure. The analysis indicates that the use of the ASTRA Bridge resulted in reductions in duration and costs (14% and 3% for the example), and a substantial reduction in user costs (51% for the example). Although more analysis is required for different types of refurbishment projects, the initial results indicate that the ASTRA Bridge may become an integral part of future highway bridge refurbishment projects. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

8. Simulation-Based Stress Tests: Integrated Simulation Approach for Assessing Resilience of Transport Systems under Multiple Stressors.

Author

Nasrazadani, Hossein, Adey, Bryan T., and Dorren, Luuk

Subjects
RAINFALL, TEST systems, LANDSLIDES, FLOOD warning systems, CLIMATE change, NATURAL disaster warning systems, CONFORMANCE testing
Abstract

A transportation system stress test is a set of one or more hypothetical scenarios designed to help determine if a transport system can continue to provide an acceptable level of service when subjected to one or more potentially disruptive events. Stress tests, if done well, enable policy makers, regulators, and network operators to assess the ability of transport systems to withstand and recover from potentially disruptive events, and modify the systems if necessary. Although work has been conducted on how to conduct high-level preliminary transportation stress tests, there has not yet been work conducted on how to conduct detailed quantitative simulation-based stress tests on transportation systems. This paper fills this gap by presenting both an approach to conduct simulation-based stress tests and example simulation-based stress tests for a transportation system subject to potentially disruptive hydrometeorological events. Three types of stress tests are conducted for the transportation system in the region of Chur, Switzerland, which is subject to heavy rainfalls that may result in floods and landslides. These include the stress caused by climate change leading to increasing the intensity of extreme rainfall events in the future, stress caused by uncertain future mobility behavior of people leading to excessive travel demand on the network, and stress caused by poor planning of the required resources for restoration interventions leading to having less resources for restoration than planned. It is argued that conducting such stress tests makes it clear to decision makers whether or not a transportation system is sufficiently resilient against various stressors, and if not, gives clear indications as to how it could be improved. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

10. Generation of Spatially Embedded Random Networks to Model Complex Transportation Networks

Author

Hackl, Jürgen and Adey, Bryan T.

Subjects
Physics - Physics and Society
Abstract

Random networks are increasingly used to analyse complex transportation networks, such as airline routes, roads and rail networks. So far, this research has been focused on describing the properties of the networks with the help of random networks, often without considering their spatial properties. In this article, a methodology is proposed to create random networks conserving their spatial properties. The produced random networks are not intended to be an accurate model of the real-world network being investigated, but are to be used to gain insight into the functioning of the network taking into consideration its spatial properties, which has potential to be useful in many types of analysis, e.g. estimating the network related risk. The proposed methodology combines a spatial non-homogeneous point process for vertex creation, which accounts for the spatial distribution of vertices, considering clustering effects of the network and a hybrid connection model for the edge creation. To illustrate the ability of the proposed methodology to be used to gain insight into a real world network, it is used to estimate standard structural statistics for part of the Swiss road network, and these are then compared with the known values., Comment: 7 pages, 4 figures, 2 tables

Published
2016
Full Text
View/download PDF

23. Automated early estimation of bridge interventions, possession windows and costs

Author

Mehranfar, Hamed, Adey, Bryan T, Moghtadernejad, Saviz, and Chuo, Steven

Abstract

Bridge managers need to generate a complete overview of required interventions, possession windows and likely costs 10–20 years ahead of execution. These, even if approximate, help ensure stable train schedules. With the increasing amount of available data and the increasing desire to exploit digitalisation to improve decision making, bridge managers are perfectly poised to make or improve these estimates by moving on from current qualitative methods. This paper proposes a way in the current climate to use digitalisation and existing data to generate approximate overviews of required interventions 10–20 years ahead of time, including estimates of component-level bridge interventions, possession windows, likely costs and the increases in failure risks if interventions are postponed. A demonstration is done on 41 bridges of a 25 km railway network in Switzerland. It is argued that the algorithm generates a more complete and consistent overview of component-level interventions, possession windows and costs compared with current qualitative methods. Additionally, the algorithm generates a solid basis for the initiation of detailed investigations of the bridges by engineering offices – that is, the investigations that result in the information required for scheduling an intervention, as well as estimating the type of intervention and track possession that are required.

Published
2024
Full Text
View/download PDF

30. Evaluating responsive gateways amid future uncertainty at the London Bridge station (UK)

Author

Lai, Bryant M L, Martani, Claudio, Roman, Orlando, and Adey, Bryan T

Abstract

The high concentration of people in train stations during rush hour poses a significant risk to passenger comfort and safety. Recent technological advancements in real-time crowd counting and pedestrian microsimulations have made it feasible to create responsive gateways for managing passenger flow to prevent overcrowding situations. This paper proposes a simulation-based cost–benefit analysis, utilising the real options methodology commonly employed to assess flexible infrastructure, to estimate the costs and benefits associated with implementing responsive gateways at the London Bridge station, UK, while considering uncertainties in passenger flow and the interests of various stakeholders. Results from the demonstrative example indicate that responsive gateways offer a better service than static ones mainly due to the expected reduction in the risks related to safety, travel time and comfort, which compensates for a higher cost of installation and maintenance. Further research is envisioned to enhance the modelling of passenger flow uncertainties and expand the objective function to encompass additional externalities. The simulation-based cost–benefit analysis proposed in this paper can be extended to a broader range of intelligent transportation systems, considering uncertainties and their diverse impacts on society. Thus, this paper presents a tool that contributes to a more transparent and objective investment decision making process, particularly concerning the development of smart cities.

Published
2024
Full Text
View/download PDF

39. A network model to optimally group road maintenance interventions for work zone development.

Author

Xu, Zhepu, Burkhalter, Marcel, Adey, Bryan T., and Yang, Qun

Subjects
ROAD work zones, OPERATIONS research
Abstract

One of the main tasks of road managers is to plan maintenance interventions, which is currently in practice often done using a simple ranking-based method. However, an increasing number of researchers are focusing their efforts on supporting the planning process using operations research models. Models that have been developed can optimally group maintenance interventions: (1) in a single work zone, taking into consideration both work zone and asset related costs, and (2) in multiple work zones, taking into consideration only asset-related costs. In this paper, a new operation research model is developed to optimally group maintenance interventions in multiple work zones, taking into consideration both work zone and asset-related costs. The new model is used to optimally group maintenance interventions for a road network located in Fujian province in China. The improvement obtained from using the new model is compared to the result determined using an operation research model that can only determine one optimal group of intervention at a time, and one determined using a commonly used ranking-based method. It is shown that the new model results in the highest net-benefit, and that both operation research models result in higher net benefits than the ranking-based method. [ABSTRACT FROM AUTHOR]

Published
2023
Full Text
View/download PDF

44. Wider view over bicycle accidents: Complementing and extending bicycle accident statistics in urban areas using surveys

Author

Ringel, Laura, Kielhauser, Clemens, Adey, Bryan T., Ringel, Laura, Kielhauser, Clemens, and Adey, Bryan T.

Abstract

City traffic planners are striving to adapt their infrastructure to not only increase the number of cyclists but also to ensure that city cycling is both enjoyable and safe. In Switzerland and in many other countries, it is suspected that only one of ten bicycle accidents is reported to the police [ 1-5]. Only knowing about 10% of the accidents, on top of the fact that there are luckily not many accidents from a statistical perspective, casts doubt about where efforts should be made to improve cycling infrastructure, and how effective the actions taken actually are. To deal with this lack of data, this paper proposes to use surveys of cyclists besides police records to obtain a more complete picture of the number and location of cycling accidents, including the ones not reported to police, and the locations that cyclists perceive as dangerous. The combination of survey and police reported data gives a considerably different and more complete impression of where there is potential to improve cycling infrastructure, when compared to that obtained using only police reported accidents. This work expounds how the survey responses about hazard perception and unreported accidents help provide a more complete overview of the accident potential of the existing cycling network and how they form a base of immensely useful inputs for planning improvements. [From: Introduction]

Published
2022

45. A preliminary investigation of the potential benefits of using the ASTRA Bridge for short-span bridge deck refurbishment projects in Switzerland

Author

Zumstein, Marco (author), Chen, Qian (author), Adey, Bryan T. (author), Hall, Daniel M. (author), Zumstein, Marco (author), Chen, Qian (author), Adey, Bryan T. (author), and Hall, Daniel M. (author)

Abstract

How bridge refurbishment projects are performed requires a trade-off between the speed and cost of the project and the amount of traffic disturbances during the project. A possible way to help reach a better balance between these two extremes is the ASTRA Bridge developed in Switzerland. The ASTRA Bridge is a 236-meter long steel ramp system on wheels, which is placed on top of the bridge deck undergoing refurbishment to enable vehicles to continue to pass over the bridge while construction work progresses underneath. This study illustrates new refurbishment processes by using the ASTRA Bridge and presents the first quantitative analysis of the effects of using the ASTRA Bridge on the time, costs and traffic disturbances associated with bridge refurbishment. The bridge investigated is a short-span (50 m long) highway bridge requiring refurbishment of its superstructure. The analysis indicates that the use of the ASTRA Bridge resulted in reductions in duration and costs (14% and 3% for the example), and a substantial reduction in user costs (51% for the example). Although more analysis is required for different types of refurbishment projects, the initial results indicate that the ASTRA Bridge may become an integral part of future highway bridge refurbishment projects., Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public., Design & Construction Management

Published
2022
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results