Estimation of Markovian transition probabilities for pavement deterioration forecasting

Kobayashi, Kiyoshi; Do, Myungsik; Han, Daeseok

doi:10.1007/s12205-010-0343-x

Cited by 82 publications

(48 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In particular, in [28], condition state values are randomly generated to represent a range of condition states at each ten-year interval using Weibull distribution and a Latin hypercube simulation. However the degradation pattern comes from knowledge of the specific area of concern or is somewhat assumed a priori like in [17] where a hazard exponential model is used to derive the Markov transition probabilities. While almost the entire literature encourages the use of either the two methodologies mentioned, there is a scarcity of models investigating the case where very limited field data are to be used.…”

Section: Introductionmentioning

confidence: 99%

A Two‐Dimension Dynamic Bayesian Network for Large‐Scale Degradation Modeling with an Application to a Bridges Network

Torre

Yeung

Morales-Nápoles

et al. 2017

Computer aided Civil Eng

View full text Add to dashboard Cite

Modelling the stochastic evolution of a largescale fleet or network generally proves to be challenging due to the large number of variables and their interactions. This difficulty may be compounded through complex relationships between various assets in the network. Although a great number of probabilistic graph-based models (e.g., Bayesian networks) have been developed recently to describe the behaviour of single assets, one can find significantly fewer approaches addressing a fully integrated network. An extension to the standard dynamic Bayesian network is proposed by introducing an additional dimension for multiple elements. These elements are then linked through a set of covariates which translate the probabilistic dependencies. A Markov chain is utilized to model the elements and develop a distribution-free mathematical framework to parametrize the transition probabilities without previous data. This is achieved by borrowing from Cooke's method for structured expert judgement and also applied to the quantification of the covariate relationships. Some metrics are also presented for evaluating the sensitivity of information inserted into the covariate-DBN where the focus is given on two specific type of configurations. The model is applied to a real-world example of steel bridge network in the Netherlands which are related through traffic load patterns. Numerical examples highlight the inference mechanism and show the sensitivity of information inserted in various ways. It is shown that information is most valuable very early and decreases substantially over time. Resulting observations entail the reduction of inference combinations and by extension a computational gain to select the most sensitive pieces of information.

show abstract

Section: Introductionmentioning

confidence: 99%

A Two‐Dimension Dynamic Bayesian Network for Large‐Scale Degradation Modeling with an Application to a Bridges Network

Torre

Yeung

Morales-Nápoles

et al. 2017

Computer aided Civil Eng

View full text Add to dashboard Cite

show abstract

“…Although most pavement experts or researchers already understand the importance of this, the task is never easy due to insufficient data for statistical methods that usually demand a large amount of inspection data to draw characteristics of the deterioration process of their road network. Many references noted that it is necessary to collect several thousand samples of inspection data in order to assure a high accuracy in deterioration forecasting (Sugisaki et al, 2006;Tsuda et al, 2006a). In reality, it is, however, not easy to build such a rich database for many reasons.…”

Section: Introductionmentioning

confidence: 99%

Application of Bayesian estimation method with Markov hazard model to improve deterioration forecasts for infrastructure asset management

2014

Self Cite

View full text Add to dashboard Cite

The heart of asset management systems for road infrastructure is the deterioration forecasting model. It provides the most fundamental information for better asset management. So far, there are many practices to build a reliable forecasting model using inspection data in conjunction with statistical theories. In many applications, however, an inadequate stock of inspection data or difficulty in applying sophisticated statistical methods have often been serious obstacles. As a solution, this paper suggests applying the Bayesian estimation method with a multi-state exponential hazard Markov chain model for simple and reliable deterioration forecasting for infrastructure. The main contents of this paper are an introduction of the model's framework combining Markov chain, hazard theory, and Bayesian estimation method, and a demonstration of its practical application with an empirical study. The empirical study was conducted with time-series inspection data of pavement from the Korean National Highways. The estimation results from the suggested method would be useful for improving the current pavement maintenance strategy for Korean National Highways. However, the most important message of this paper is that the framework of the Bayesian Markov hazard model could be the best model to use for other civil infrastructure that has gradual changes in condition. The great advantages from the Bayesian estimation method may facilitate development of customized asset management systems.

show abstract

“…The Markov chain has been used for the infrastructure deterioration predictions, including bridges (Fu and Debraj, 2008;Mishalani and Madanat, 2002;Ranjith, et al 2011), roads (Abaza and Ashur, 1999;Anderson, et al 1994;Butt, et al 1994;Hudson, et al 1998;Jin, et al 2010;Kobayashi, et al 2010;Lethanh and Adey, 2012;Li, 1997;Mandiartha, 2010;Mills, 2010;Ortiz-Garcia, et al 2006;Panthi, 2009;Pierce, 2003;Robinson, et al 1998;Tack and Chou, 2002;Wang, et al 1994), waste water (Hyeon-shik, et al 2006), rail (Ferreira and Murray, 1997;Shafahi and Hakhamaneshi, 2009) and pipelines (Jin, et al 2010;Sinha and Mark, 2004). …”

Section: Markov Chain Reviewmentioning

confidence: 99%

“…Eguchi, et al (2006) developed the condition index and rutting deterioration models with the help of the multistage hazard model for a homogeneous TPM. Kobayashi et al (2010) used an exponential hazard model to derive a non-homogeneous TPM. The maximum likelihood method was undertaken using the pavement structural strength and loading data to determine parameters of the exponential hazard model.…”

Section: Multi-stage Hazard Modelmentioning

confidence: 99%

See 1 more Smart Citation

Improvement of a Pavement Management System Incorporating Flooding

Khan¹

View full text Add to dashboard Cite

Natural disasters can cause severe damage to infrastructure such as the road network. Currently, a Pavement Management System (PMS) does not incorporate flooding in a LifeCycle Analysis (LCA). A few Road Deterioration (RD) models have addressed flooding, but they have limitations. As a result, there are not any comprehensive RD models that can incorporate flooding in pavement performances. In addition, no optimum life-cycle road maintenance strategy is available. No study investigated the pavement performances because of a loss in Modulus of Resilience (Mr) at granular and subgrade layers during extreme moisture intrusion. The derivation of pre-and post-flood road maintenance strategies and a flood risk assessment should also be incorporated in a PMS.As a case study, this research has considered the January 2011 flood of Queensland, Australia, and has used 34,000 km road database of the Queensland's main roads authority.The major objectives of this study are to derive: i) network and project level roughness and rutting-based RD models with flooding, ii) pavement performances due to Mr loss at granular and subgrade layers, iii) flood-resilient pavements, iv) optimum road maintenance strategies at without flood, pre-and post-flood scenarios, and v) pavements' flood risks. The current scope covered the pavements that are affected after a flood, but are not washed away completely and need rehabilitation for structural strengthening.This research has used a probabilistic approach for deriving the RD models, which are valid at both network and project levels. Moreover, the proposed RD models can estimate road deterioration after a flood at different probabilities of flooding.The actual roughness and rutting vs. time data are assessed for the representative road groups or site specific roads to get the transition probability matrices for with and without flooding conditions, which are used in a Monte Carlo simulation. The new RD models show significant pavement deterioration at different probabilities of flooding events. The results are found valid with actual data for about 2 to 3 years after the January 2011 flood. A t-test also supports this match. A pavement's performance due to Mr losses at granular layers is checked using the two renowned roughness models, which results are found close match with the actual after flood data and RD models.

show abstract

Estimation of Markovian transition probabilities for pavement deterioration forecasting

Cited by 82 publications

References 20 publications

A Two‐Dimension Dynamic Bayesian Network for Large‐Scale Degradation Modeling with an Application to a Bridges Network

A Two‐Dimension Dynamic Bayesian Network for Large‐Scale Degradation Modeling with an Application to a Bridges Network

Application of Bayesian estimation method with Markov hazard model to improve deterioration forecasts for infrastructure asset management

Improvement of a Pavement Management System Incorporating Flooding

Contact Info

Product

Resources

About