Data-driven optimization of railway maintenance for track geometry

Railway big data technologies are transforming the existing track inspection and maintenance policy deployed for railroads in North America. This paper develops a data-driven condition-based policy for the inspection and maintenance of track geometry. Both preventive maintenance and spot corrective maintenance are taken into account in the investigation of a 33-month inspection dataset that contains a variety of geometry measurements for every foot of track. First, this study separates the data based on the time interval of the inspection run, calculates the aggregate track quality index (TQI) for each track section, and predicts the track spot geo-defect occurrence probability using random forests. Then, a Markov chain is built to model aggregated track deterioration, and the spot geo-defects are modeled by a Bernoulli process. Finally, a Markov decision process (MDP) is developed for track maintenance decision making, and it is optimized by using a value iteration algorithm. Compared with the existing maintenance policy using Markov chain Monte Carlo (MCMC) simulation, the maintenance policy developed in this paper results in an approximately 10% savings in the total maintenance costs for every 1 mile of track.     

Comparison of pattern recognition techniques for the classification of impact acoustic emissions

Current day condition monitoring applications involving wood are mostly carried out through visual inspection and if necessary some impact acoustic examination is carried out. These inspections are mainly done intuitively by skilled personnel. In this paper, a pattern recognition approach has been considered to automate such intuitive human skills for the development of robust and reliable methods within the area. The study presents a comparison of several pattern recognition techniques combined with various stationary feature extraction techniques for classification of impact acoustic emissions. Further issues concerning feature fusion are discussed as well. It is hoped that this kind of broad analysis could be used to handle a wide spectrum of tasks within the area, and would provide a perfect ground for future research directions. A brief introduction to the techniques is provided for the benefit of the readers unfamiliar with the techniques.Pattern classifiers such as support vector machines, etc. are combined with stationary feature extraction techniques such as linear predictive cepstral coefficients, etc. Results from support vector machines in combination with linear predictive cepstral coefficients delivered good classification rates. However, Gaussian mixture models delivered higher classification rates when feature fusion is proposed.     

Comprehensive Review of the Maritime Safety Regimes: Present Status and Recommendations for Improvements

Abstract

This article presents a comprehensive review of the maritime safety regimes and provides recommendations on how to improve the system. The results show a complex legal framework which generates a high amount of inspections and overlapping of inspection areas where no cross‐recognition is established by the various stakeholders. While the safety system seems to be successful in eliminating substandard vessels and while average insurance claims costs are substantially lower for inspected vessels than non‐inspected vessels, the results indicate that the economic conditions of the shipping market also have an effect on safety quality besides the frequency of inspections. No significant differences can be found between industry inspections and port state control inspections with respect to decreasing the probability of casualty. The system could be made more effective by combining data sources on inspections and using them respectively to improve risk profiling and to decrease the frequency of inspections performed on ship types such as tankers. The results further indicate a lack of proper implementation of the International Safety Management Code (ISM code) and conventions with reference to working and living conditions of crew (ILO 147). A revision of the ISM code and more emphasis on enforcement of ILO 147 could further enhance the level of safety at sea. The authors would like to thank several inspection regimes for their cooperation in providing inspection data and in allowing the observation of surveys and inspections on 26 vessels. In addition, the authors would like to acknowledge the data providers for the casualty data, Clarksons for the economic data as well as two P&I Clubs in making data on insurance claims available.     

Principal components analysis and track quality index: A machine learning approach

Track geometry data exhibits classical big data attributes: value, volume, velocity, veracity and variety. Track Quality Indices-TQI are used to obtain average-based assessment of track segments and schedule track maintenance. TQI is expressed in terms of track parameters like gage, cross-level, etc. Though each of these parameters is objectively important but understanding what they collectively convey for a given track segment often becomes challenging. Several railways including passenger and freight have developed single indices that combines different track parameters to assess overall track quality. Some of these railways have selected certain parameters whilst dropping others. Using track geometry data from a sample mile track, we demonstrate how to combine track geometry parameters into a low dimensional form (TQI) that simplifies the track properties without losing much variability in the data. This led us to principal components. To validate the use of principal components as TQI, we employed a two-phase approach. First phase was to identify a classic machine learning technique that works well with track geometry data. The second step was to train the identified machine learning technique on the sample mile-track data using combined TQIs and principal components as defect predictors. The performance of the predictors were compared using true and false positive rates. The results show that three principal components were better at predicting defects and revealing salient characteristics in track geometry data than combined TQIs even though there were some correlations that are potentially useful for track maintenance.     

Failure rates and data driven policies for vehicle safety inspections in Pennsylvania

Rail, truck, commercial bus, and aircraft have federally mandated safety inspection programs in the United States, while inspections of personal vehicles, which make up the majority of passenger miles, are optionally imposed at the state level. In recent years, some states have chosen to eliminate the vehicle safety inspection program because of budget constraints and concerns about program effectiveness. Currently, 26 states have a schedule for conducting safety inspections, but Pennsylvania is one of thirteen states that currently require all personal light duty vehicles to be inspected every year. The remaining states have completely eliminated safety inspection programs. However, as automobiles become safer, Pennsylvania legislators are now pushing to phase out the inspection program to reduce the costs of owning a vehicle. This study combines Pennsylvania vehicle registration data with two large samples of results from state safety inspections. We find that the state safety inspection fail rate for light-duty vehicles is 12–18%, well above the often-cited rate of 2%. Vehicles that are older than three years old or have more than about 30,000 miles can have much higher rates. When analyzing new vehicles, less than or equal to one year old, it is found that even these vehicles have a failure rate greater than zero. Furthermore, while the vehicle fleet appears to be getting safer over the past few years by improvements in technology or other external circumstances, the inspection failure rate does not appear to be trending toward zero in the near future. We also show that accurate inspection data is limited and often incorrectly analyzed. Lastly, the importance of vehicle maintenance over a vehicle’s lifetime is proven to be evident, since regular usage causes vehicles to deteriorate. We conclude that vehicle safety inspections should continue to be implemented in order to keep driving conditions safe.     

漏磁检测技术在管道检测中的应用及影响因素分析

阐述了漏磁检测的原理,介绍了基于漏磁原理的检测系统组成,以及在长输管道及工业管道检测中的工程应用。详细分析了漏磁检测技术的主要影响因素。指出国内漏磁检测技术领域与国外存在较大差距。国内管道内检测已进入立法阶段,相关标准的初稿已基本完成,未来漏磁检测技术将在维护管道安全生产上发挥越来越重要的作用。     

Condition monitoring of wooden railway sleepers

Wooden railway sleeper inspections in Sweden are currently performed manually by a human operator; such inspections are to large extent based on visual analysis. In this paper a machine vision based approach has been considered to emulate the visual abilities of the human operator to enable automation of the process. Digital images from either ends (left and right) of the sleepers have been acquired. A pattern recognition approach has been adopted to classify the condition of the sleeper into classes (good or bad) and thereby achieve automation. Appropriate image analysis techniques were applied and relevant features such as the number of cracks on a sleeper, average length and width of the crack and the condition of the metal plate were determined. Feature fusion has been proposed in order to integrate the features obtained from each end for the classification task which follows. The effect of using classifiers like multi-layer perceptron and support vector machines has been tested and compared. Results obtained from the experiments show that multi-layer perceptron and support vector machines have achieved encouraging results, with a classification accuracy of 90%; thereby exhibiting a competitive performance when compared to a human operator.     

管道环焊缝X射线数字成像动态采集软件设计

目前,国内的管道环焊缝数字射线检测设备仅能实现静态成像,曝光时间过长,成像结果由多张图像组成,增加了判读工作量和图像评定难度。管道环焊缝X射线数字成像动态采集系统解决了这些问题,采集面板在电机的驱动下沿着轨道匀速行驶,采集图像无缝无重叠地连续实时显示,最终形成一幅完整的管道焊缝扫查图。针对数字射线动态数据采集系统的TDI原理、TDI扫查的时钟同步源、CMOS平板探测器中偏置电压的刷新和面板校准等问题进行详细说明,完成了X射线数字成像动态采集软件的设计和编程,并在管径813 mm、壁厚12.5 mm的管道环焊缝上进行检测试验,采集图像中缺陷清晰可见。     

Map-matching poor-quality GPS data in urban environments: the pgMapMatch package

Global Positioning System (GPS) data have become ubiquitous in many areas of transportation planning and research. The usefulness of GPS data often depends on the points being matched to the true sequence of edges on the underlying street network – a process known as ‘map matching.’ This paper presents a new map-matching algorithm that is designed for use with poor-quality GPS traces in urban environments, where drivers may circle for parking and GPS quality may be affected by underground parking and tall buildings. The paper is accompanied by open-source Python code that is designed to work with a PostGIS spatial database. In a test dataset that includes many poor-quality traces, our new algorithm accurately matches about one-third more traces than a widely available alternative. Our algorithm also provides a ‘match score’ that evaluates the likelihood that the match for an individual trace is correct, reducing the need for manual inspection.     

Policing the poor: The impact of vehicle emissions inspection programs across income

Comparing vehicle emissions inspection results with vehicle owner income shows that the Arizona vehicle emissions inspection program constrains the vehicle repair decisions of people in the low end of the income distribution more than people in the high end. Individuals who live in areas with lower annual income are both (i) more likely to drive vehicles that fail emissions inspections at a higher average rate, and (ii) more likely to fail emissions inspections conditional on vehicle characteristics. The top income quintile fails emissions inspections 20% less often than the bottom income quintile even when controlling for observable vehicle characteristics. This implies that owner characteristics, in addition to observable vehicle characteristics, have a non-negligible impact on vehicle emissions rates. Therefore, the impact of programs designed to reduce vehicle emissions could be greater if participation were subject to a means test.     

Open PFLOW: Creation and evaluation of an open dataset for typical people mass movement in urban areas

Understanding people flow at a citywide level is critical for urban planning and commercial development. Thanks to the ubiquity of human location tracking devices, many studies on people mass movement with mobility logs have been conducted. However, high cost and severe privacy policy constraints still complicate utilization of these data in practice. There is no dataset that anyone can freely access, use, modify, and share for any purpose. To tackle this problem, we propose a novel dataset creation approach (called Open PFLOW) that continuously reports the spatiotemporal positions of all individual’s in urban areas based on open data. With fully consideration of the privacy protection, each entity in our dataset does not match the actual movement of any real person, so that the dataset can be totally open to public as part of data infrastructure. Because the result is shown at a disaggregate level, users can freely modify, process, and visualize the dataset for any purpose. We evaluate the accuracy of the dataset by comparing it with commercial datasets and traffic census indicates that it has a high correlation with mesh population and link-based traffic volume.     

Second-order models and traffic data from mobile sensors

Mobile sensing enabled by GPS or smart phones has become an increasingly important source of traffic data. For sufficient coverage of the traffic stream, it is important to maintain a reasonable penetration rate of probe vehicles. From the standpoint of capturing higher-order traffic quantities such as acceleration/deceleration, emission and fuel consumption rates, it is desirable to examine the impact on the estimation accuracy of sampling frequency on vehicle position. Of the two issues raised above, the latter is rarely studied in the literature. This paper addresses the impact of both sampling frequency and penetration rate on mobile sensing of highway traffic. To capture inhomogeneous driving conditions and deviation of traffic from the equilibrium state, we employ the second-order phase transition model (PTM). Several data fusion schemes that incorporate vehicle trajectory data into the PTM are proposed. And, a case study of the NGSIM dataset is presented which shows the estimation results of various Eulerian and Lagrangian traffic quantities. The findings show that while first-order traffic quantities can be accurately estimated even with a low sampling frequency, higher-order traffic quantities, such as acceleration, deviation, and emission rate, tend to be misinterpreted due to insufficiently sampled vehicle locations. We also show that a correction factor approach has the potential to reduce the sensing error arising from low sampling frequency and penetration rate, making the estimation of higher-order quantities more robust against insufficient data coverage of the highway traffic.     

Probabilistic risk analysis of flying ballast hazard on high-speed rail lines

Flying ballast is a significant safety concern for high-speed train operations on ballasted tracks. It is the phenomenon of a ballast particle displaced from the track, due to the aerodynamic force induced by a passing train traveling above a certain speed. Flying ballast can potentially damage tracks and rolling stock, thereby posing a risk to high-speed rail operations. This paper develops a Probabilistic Risk Analysis (PRA) model based on the information available from the field and the literature. The model enables a quantitative assessment of the probability of ballast particle displacement at a particular position on the track, as well as the probabilistic distribution of the total number of ballast particles that are expected to move. The model accounts for various risk factors, such as train speed, ballast gradation, and track position. The model application is illustrated using a ballasted track on the Yellow River Bridge on the Beijing-Shanghai high-speed rail line in China. The analysis finds that flying ballast probability increases when train speed increases, in particular, the problem of flying ballast becomes more pronounced when train speed exceeds 350 km per hour (217 miles per hour). Flying ballast probability might be reduced when the ballast profile is lower, given all else being equal. In addition, flying ballast probability is expected to be higher at the center of the track than in other positions. The proposed risk model can be further developed and ultimately be used to evaluate route-specific flying ballast risk, enabling the identification, assessment, and comparison of risk mitigation strategies in order to support emerging high-speed rail operations.     

Improving rail network velocity: A machine learning approach to predictive maintenance

Rail network velocity is defined as system-wide average speed of line-haul movement between terminals. To accommodate increased service demand and load on rail networks, increase in network velocity, without compromising safety, is required. Among many determinants of overall network velocity, a key driver is service interruption, including lowered operating speed due to track/train condition and delays caused by derailments. Railroads have put significant infrastructure and inspection programs in place to avoid service interruptions. One of the key measures is an extensive network of wayside mechanical condition detectors (temperature, strain, vision, infrared, weight, impact, etc.) that monitor the rolling-stock as it passes by. The detectors are designed to alert for conditions that either violate regulations set by governmental rail safety agencies or deteriorating rolling-stock conditions as determined by the railroad.Using huge volumes of historical detector data, in combination with failure data, maintenance action data, inspection schedule data, train type data and weather data, we are exploring several analytical approaches including, correlation analysis, causal analysis, time series analysis and machine learning techniques to automatically learn rules and build failure prediction models. These models will be applied against both historical and real-time data to predict conditions leading to failure in the future, thus avoiding service interruptions and increasing network velocity. Additionally, the analytics and models can also be used for detecting root cause of several failure modes and wear rate of components, which, while do not directly address network velocity, can be proactively used by maintenance organizations to optimize trade-offs related to maintenance schedule, costs and shop capacity. As part of our effort, we explore several avenues to machine learning techniques including distributed learning and hierarchical analytical approaches.     

An autonomous system for maintenance scheduling data-rich complex infrastructure: Fusing the railways’ condition,planning and cost

National railways are typically large and complex systems. Their network infrastructure usually includes extended track sections, bridges, stations and other supporting assets. In recent years, railways have also become a data-rich environment.Railway infrastructure assets have a very long life, but inherently degrade. Interventions are necessary but they can cause lateness, damage and hazards. Every day, thousands of discrete maintenance jobs are scheduled according to time and urgency. Service disruption has a direct economic impact. Planning for maintenance can be complex, expensive and uncertain.Autonomous scheduling of maintenance jobs is essential. The design strategy of a novel integrated system for automatic job scheduling is presented; from concept formulation to the examination of the data to information transitional level interface, and at the decision making level. The underlying architecture configures high-level fusion of technical and business drivers; scheduling optimized intervention plans that factor-in cost impact and added value.A proof of concept demonstrator was developed to validate the system principle and to test algorithm functionality. It employs a dashboard for visualization of the system response and to present key information. Real track incident and inspection datasets were analyzed to raise degradation alarms that initiate the automatic scheduling of maintenance tasks. Optimum scheduling was realized through data analytics and job sequencing heuristic and genetic algorithms, taking into account specific cost & value inputs from comprehensive task cost modelling. Formal face validation was conducted with railway infrastructure specialists and stakeholders. The demonstrator structure was found fit for purpose with logical component relationships, offering further scope for research and commercial exploitation.     

Cellpath: Fusion of cellular and traffic sensor data for route flow estimation via convex optimization

A new convex optimization framework is developed for the route flow estimation problem from the fusion of vehicle count and cellular network data. The issue of highly underdetermined link flow based methods in transportation networks is investigated, then solved using the proposed concept of cellpaths for cellular network data. With this data-driven approach, our proposed approach is versatile: it is compatible with other data sources, and it is model agnostic and thus compatible with user equilibrium, system-optimum, Stackelberg concepts, and other models. Using a dimensionality reduction scheme, we design a projected gradient algorithm suitable for the proposed route flow estimation problem. The algorithm solves a block isotonic regression problem in the projection step in linear time. The accuracy, computational efficiency, and versatility of the proposed approach are validated on the I-210 corridor near Los Angeles, where we achieve 90% route flow accuracy with 1033 traffic sensors and 1000 cellular towers covering a large network of highways and arterials with more than 20,000 links. In contrast to long-term land use planning applications, we demonstrate the first system to our knowledge that can produce route-level flow estimates suitable for short time horizon prediction and control applications in traffic management. Our system is open source and available for validation and extension.     

Probabilistic speed–density relationship for pedestrian traffic

We propose a probabilistic modeling approach to represent the speed–density relationship of pedestrian traffic. The approach is data-driven, and it is motivated by the presence of high scatter in the raw data that we have analyzed. We show the validity of the proposed approach, and its superiority compared to deterministic approaches from the literature using a dataset collected from a real scene and another from a controlled experiment.     

Probe vehicle based real-time traffic monitoring on urban roadways

Travel time estimation and prediction on urban arterials is an important component of Active Traffic and Demand Management Systems (ATDMS). This paper aims in using the information of GPS probes to augment less dynamic but available information describing arterial travel times. The direction followed in this paper chooses a cooperative approach in travel time estimation using static information describing arterial geometry and signal timing, semi-dynamic information of historical travel time distributions per time of day, and utilizes GPS probe information to augment and improve the latter. First, arterial travel times are classified by identifying different travel time states, then link travel time distributions are approximated using mixtures of normal distributions. If prior travel time data is available, travel time distributions can be estimated empirically. Otherwise, travel time distribution can be estimated based on signal timing and arterial geometry. Real-time GPS travel time data is then used to identify the current traffic condition based on Bayes Theorem. Moreover, these GPS data can also be used to update the parameters of the travel time distributions using a Bayesian update. The iterative update process makes the posterior distributions more and more accurate. Finally, two comprehensive case studies using the NGSIM Peachtree Street dataset, and GPS data of Washington Avenue in Minneapolis, were conducted. The first case study estimated prior travel time distributions based on signal timing and arterial geometry under different traffic conditions. Travel time data were classified and corresponding distributions were updated. In addition, results from the Bayesian update and EM algorithm were compared. The second case study first tested the methodologies based on real GPS data and showed the importance of sample size. In addition, a methodology was proposed to distinguish new traffic conditions in the second case study.     

Use of a novel dataset to explore spatial and social variations in car type,size, usage and emissions

The ’MOT’ vehicle inspection test record dataset recently released by the UK Department for Transport (DfT) provides the ability to estimate annual mileage figures for every individual light duty vehicle greater than 3 years old within Great Britain. Vehicle age, engine size and fuel type are also provided in the dataset and these allow further estimates to be made of fuel consumption, energy use, and per vehicle emissions of both air pollutants and greenhouse gases. The use of this data permits the adoption of a new vehicle-centred approach to assessing emissions and energy use in comparison to previous road-flow and national fuel consumption based approaches. The dataset also allows a spatial attribution of each vehicle to a postcode area, through the reported location of relevant vehicle testing stations. Consequently, this new vehicle data can be linked with socio-demographic data in order to determine the potential characteristics of vehicle owners.This paper provides a broad overview of the types of analyses that are made possible by these data, with a particular focus on distance driven and pollutant emissions. The intention is to demonstrate the very broad potential for this data, and to highlight where more focused analysis could be useful. The findings from the work have important implications for understanding the distributional impacts of transport related policies and targeting messaging and interventions for the reduction of car use.     

Network Signal Setting Design: Meta-heuristic optimisation methods

This paper aims to investigate the application of meta-heuristic optimisation methods to Network Signal Setting Design. The adopted approaches are (i) three step optimisation, in which first the stage matrix (stage composition and sequence), the green timings at each single junction are optimised, then the node offsets are computed in three successive steps; (ii) two step optimisation, in which the stage matrix is defined at a first step, then the green timings and the node offsets are computed at a second step. In both approaches the stage matrix optimisation is carried out through explicit complete enumeration.In the first approach multi-criteria optimisation is followed for single junction signal setting design (green timings), whilst the coordination (node offsets) is approached through mono-criterion optimisation, as well as for the synchronisation (green timings and offsets) in the second approach.A new traffic flow model mixing CTM and PDM has been applied. This model allows to explicitly represent horizontal queuing phenomena as well as dispersion along a link. Some meta-heuristic algorithms (i.e. Genetic Algorithms, Hill Climbing and Simulated Annealing) are investigated in order to solve the two problems.The proposed strategies are applied to two different layouts (a two junction arterial vs. a four junction network) and their effectiveness is evaluated by comparing the obtained results with those from benchmark approaches implementing mono-criterion optimisation only.