Ministry Of Transportation Of Ontario Research Articles

Exploitation of deep learning in the automatic detection of cracks on paved roads

With the advance of deep learning networks, their applications in the assessment of pavement conditions are gaining more attention. A convolutional neural network (CNN) is the most commonly used network in image classification. In terms of pavement assessment, most existing CNNs are designed to only distinguish between cracks and non-cracks. Few networks classify cracks in different levels of severity. Information on the severity of pavement cracks is critical for pavement repair services. In this study, the state-of-the-art CNN used in the detection of pavement cracks was improved to localize the cracks and identify their distress levels based on three categories (low, medium, and high). In addition, a fully convolutional network (FCN) was, for the first time, utilized in the detection of pavement cracks. These designed architectures were validated using the data acquired on four highways in Ontario, Canada, and compared with the ground truth that was provided by the Ministry of Transportation of Ontario (MTO). The results showed that with the improved CNN, the prediction precision on a series of test image patches were 72.9%, 73.9%, and 73.1% for cracks with the severity levels of low, medium, and high, respectively. The precision for the FCN was tested on whole pavement images, resulting in 62.8%, 63.3%, and 66.4%, respectively, for cracks with the severity levels of low, medium, and high. It is worth mentioning that the ground truth contained some uncertainties, which partially contributed to the relatively low precision.

Use of Pavement Management System Data to Enhance Pavement Performance Specifications in Canada

Most agency materials and construction specifications for pavement rehabilitation and construction activities provide little or no linkage between quality assurance methods and the in-service performance of the treatment. The Ministry of Transportation of Ontario (MTO), Canada, is developing and implementing performance specifications to bridge this gap. MTO defines performance specifications as specifications that describe how the finished product should perform over time. This paper describes how MTO used historical pavement management system data to assess performance sensitivity and to create historical performance distributions and how the agency used these distributions to set acceptance criteria, including payment adjustment for performance.

Effect of shear stud clusters in composite girder bridge design

As part of ongoing efforts to accelerate bridge construction in Ontario, the Ministry of Transportation of Ontario (MTO) has turned increasingly to prefabricated bridge technology as a bridge construction method when conditions allow. One of the prefabricated deck systems commonly used by MTO involves installing precast full-depth deck panels with pre-formed shear pockets on top of the naked steel girders, which are then made composite with the girders through shear studs installed inside the shear pockets. Construction of the deck slab is completed by filling the shear pockets with in situ concrete. Recent prefabricated bridge projects have shown an increasing tendency to concentrate the shear studs into clusters of closely-spaced studs in shear pockets that are spaced at relatively large distances apart. Concerns have been raised about the effectiveness of the resulting composite action between the precast panels and the supporting girders. As a result, MTO’s Bridge Office recently carried out a comprehensive experimental research study to investigate the issue by using reduced scale models of single composite beam systems as well as shear push-out specimens. This paper describes the laboratory research project and the tests that were carried out and discusses the results obtained. Comparisons with theoretical results are made to assess the composite action that is developed. Experimental test results indicate that closely-spaced studs in the shear pockets in the precast deck panels provide adequate composite action in composite bridge girder design.

Development of regression equations for local calibration of rutting and IRI as predicted by the MEPDG models for flexible pavements using Ontario's long-term PMS data

Local calibration is an important step before a transportation agency adopts the American Association of State Highway and Transportation Officials' (AASHTO) mechanistic-empirical pavement design guide (MEPDG). This paper presents the challenges of and findings from the local calibration of flexible pavements in provincial highways under the jurisdiction of the Ministry of Transportation of Ontario (MTO). A calibration database was developed that involved a hierarchical framework of the input parameters required for AASHTOWare Pavement ME (the MEPDG software) and the historical field performance data based on the MTO's second-generation pavement management system. A regression analysis is carried out for preliminary calibration of rutting and international roughness index (IRI) models by comparing the predicted distress to observed distress. The analysis suggested that whereas the MEPDG provided fairly unbiased prediction of the IRI value, it often over-predicted the total rutting. Calibrated predicted IRI and rut depth are found for Ontario's local conditions from MEPDG distress prediction models. A further clustering analysis based on Functional Class and geographical zone for the rutting and IRI, respectively, improved the precision of the locally calibrated models.

Canadian calibration on Mechanistic - Empirical Pavement Design Guide to estimate International Roughness Index (IRI) using MTO data

This paper presents flexible pavement performance models developed for the Ministry of Transportation of Ontario (MTO) by using data from the MTO's Pavement Management System (PMS2). The performance model coefficients have been developed for application in the Mechanistic - Empirical Pavement Design Guide (MEPDG) and were calibrated using statistical tools through a series of analyses on historical pavement condition data that was collected in the field. The statistical analysis involved collection of historical data and development of pavement model categories. It was then classified according to pavement type, equivalent total pavement thickness, traffic volume, soil type, and climatic zone. In the development of the performance curves, 75% of the data was used to calibrate the performance curves, which is described by the predicted Pavement Condition Index (PCI) as a function of pavement age in years. The remaining 25% of the data was used to validate the various performance models using various statistical tools. The procedure and analysis methodology used in the development of the performance models are presented in the paper.The paper provides a practical framework for comparing existing PMS2 flexible performance curves to performance predictions obtained from the MEPDG. Example case studies for typical Ontario roads are presented in the paper in terms of statistical analysis method.

Renewal and Update of MTO IDF Curves: Defining the Uncertainty

The University of Waterloo has been commissioned by the Design and Contract Standards Office of the Ministry of Transportation of Ontario (MTO) to update the i…

Evaluation of Network RTK Performance and Elements of Certification—A Southern Ontario Case Study

Over the past decade, network RTK technology has become popular as an efficient method of precise, real-time positioning. Its relatively low-cost and single receiver ease-of-use has allowed it to mostly replace static relative GPS and single baseline RTK in urban areas where such networks are economically viable (e.g., cadastral and construction survey). The Ministry of Transportation of Ontario (MTO) and York University have investigated the performance of commercial network RTK services in Southern Ontario, where performance is defined by a set of developed metrics. It was found that the user horizontal solution had an overall precision of ∼2.5 cm (95%), though there were cases of solution biases, drifts and gaps. A follow-up study is developing criteria and pathways for the certification of such commercial network RTK services, focusing on: reference station integration, reference station maintenance, and user solution monitoring. A set of recommendations for network certification is in preparation.

A Stochastic Model for Highway Accident Predictions with Winter Data

In this paper, we consider the problem of modeling and predicting highway accidents in the presence of randomly changing winter driving conditions. Unlike most accident prediction models in the literature, which are typically formulated in a static (e.g. regression models) or discrete time (e.g. time-series models) setting, we propose a continuous- time stochastic model to describe the relation between highway accidents and winter weather dynamics. We believe this to be a more natural way to describe discrete-event highway accidents that occur in continuous-time. In particular, the accident counting process is viewed as a non-homogeneous Poisson process (NHPP) with an intensity function that depends on a (Markovian) weather process. Such a model is known in the stochastic process literature as a Markov- modulated Poisson process (MMPP) and has been successfully applied to queuing and telecommunications problems. One main advantage of such an approach, is its ability to provide explicit closed-form prediction formulae for both weather and accidents over any future time horizon (i.e. short or long-term predictions). To illustrate the effectiveness of the proposed stochastic model, we study a large winter data set provided by Ministry of Transportation of Ontario (MTO) that includes motor vehicle accidents on Highway 401, the busiest highway in North America.

Eliminating Deck Joints Using Debonded Link Slabs: Research and Field Tests in Ontario

One of the main factors affecting the durability of bridge structures is the presence of expansion joints at bridge support locations. The inability of current joint systems to provide reliable, long-term, leak-proof performance generally leads to early leakage of chloride-contaminated water through these joints, thereby causing premature corrosion in the deck elements below. This problem is particularly evident in older-type multispan bridges in which the girders are simply supported at the piers and are separated by expansion joints or simple paved-over joints. To address this problem, the Ministry of Transportation of Ontario (MTO) has recently rehabilitated a number of bridge decks using a debonded link slab system to replace the deck joints at the pier locations. To get a better understanding of the performance and reliability of this new rehabilitative technique, MTO recently carried out an experimental research study of the long-term performance of the system on scale test models that were subjected to extensive cyclic loading in the laboratory. At the same time, it carried out a load test of a recently rehabilitated structure to study its structural behavior both before and after the link slab was constructed. The test structure was instrumented with sensors that measured deflections and strains in the link slab and girders. This paper describes the experimental research study and the behavioral load tests that were carried out, and discusses the results obtained. The experimental study showed that the long-term performance of the link slab was not affected by the extensive cyclic loading to which the model was subjected, whereas the load testing of the test structure showed that it satisfied the serviceability limit state requirements of the Canadian Highway Bridge Design Code, thus validating the design methodology of the system.

The Canadian Safe Driving Study—Phase I pilot: Examining potential logistical barriers to the full cohort study

Multiple organizations and task forces have called for a reliable and valid method to identify older drivers who are medically unfit to drive. The development of a clinical decision rule for this type of screening requires data from a longitudinal prospective cohort of older drivers. The aim of this article is to identify potential design, sampling and data collection barriers to such studies based on an analysis of the Canadian Safe Driving Study—phase I pilot (Candrive I). A convenience sample of 100 active older drivers aged 70years or more was recruited through the aid of a seniors’ organization, 94 of whom completed the full study (retention rate 94%). Data were collected over the course of 1year on various driving behaviours, as well as on cognitive, physical and mental functioning. Driving patterns were recorded using driving diaries, logs and electronic devices. Driving records from the Ministry of Transportation of Ontario (MTO) were obtained for the 3-year period preceding the study initiation and up to 1year following study completion. An increased burden of illness was observed as the number of medical diagnoses and medication use increased over the study period. Study participants were involved in a total of five motor vehicle collisions identified through MTO records, which was comparable to the Ontario annual collision rate of 4.1% for drivers aged 75years or older. In sum, many of the relevant logistical and practical barriers to studying a large sample of older drivers longitudinally have been shown to be addressable, supporting the feasibility of completing a large prospective cohort study of older drivers.

Department of Transportation vs Self-reported Data on Motor Vehicle Collisions and Driving Convictions for Stroke Survivors: Do They Agree?

Objective: Research on stroke survivors’ driving safety has typically used either self-reports or government records, but the extent to which the 2 may differ is not known. We compared government records and self-reports of motor vehicle collisions and driving convictions in a sample of stroke survivors. Methods: The 56 participants were originally recruited for a prospective study on driving and community re-integration post-stroke; the study population consisted of moderately impaired stroke survivors without severe communication disorders who had been referred for a driving assessment. The driving records of the 56 participants for the 5 years before study entry and the 1-year study period were acquired with written consent from the Ministry of Transportation of Ontario (MTO), Canada. Self-reports of collisions and convictions were acquired via a semistructured interview and then compared with the MTO records. Results: Forty-three participants completed the study. For 7 (13.5%) the MTO records did not match the self-reports regarding collision involvement, and for 9 (17.3%) the MTO records did not match self-reports regarding driving convictions. The kappa coefficient for the correlation between MTO records and self-reports was 0.52 for collisions and 0.47 for convictions (both in the moderate range of agreement). When both sources of data were consulted, up to 56 percent more accidents and up to 46 percent more convictions were identified in the study population in the 5 years before study entry compared to when either source was used alone. Conclusion: In our population of stroke survivors, self-reports of motor vehicle collisions and driving convictions differed from government records. In future studies, the use of both government and self-reported data would ensure a more accurate picture of driving safety post-stroke.

Alternative modeling framework for pervious concrete pavement condition analysis

Pervious concrete pavement (PCP) is listed as one of the best stormwater management practices by the Environmental Protection Agency (EPA) due to its capacity to reduce impervious areas, capability to improve water quality, reduce runoff quantity, eliminate the need for detention basins, protect downstream channels, and reduce flooding (save lives and properties). Moreover, it is designated a “Green Asset” since it allows stormwater to seep through pavement to recharge ground water level. PCP may potentially improve skid resistance, reduce noise, decrease heat, and reduce water contamination. However, since PCP must have adequate permeability (porous structure), it can potentially be susceptible to freeze–thaw damage particularly in cold climates. To deploy the PCP in a transportation network, its performance should be thoroughly evaluated. For this purpose, the condition of PCP has to be defined and determined over time. Since PCP has not been widely investigated in cold climates such as Canada, no condition indices have been defined for it. This paper proposes a combined condition index for PCP to ensure that it will perform adequately over time. To develop a combined condition index, both a panel rating approach and objective measurements are employed to assess surface distresses and permeability rates of specific PCP sections. The panel rating method is applied due to the lack of long-term performance data to evaluate PCP sections’ conditions with regards to surface distress (e.g., raveling, spalling, and cracking) and functional performance (i.e., permeability rate). The objective measurements of surface distress and permeability rate (functional performance) of the same PCP sections are performed through the application of the adjusted Ministry of Transportation of Ontario (MTO) protocol and a permeameter, respectively. Regression analysis is used to relate surface objective measurements to the surface panel rating. As a result of statistical analyses, the panel rating method is successfully applied only in surface distress rating. Accordingly, a combined condition index is developed using surface distress ratings and objective functional performance measures.

Validation and Implementation of Ontario, Canada, Network-Level Distress Guidelines and Condition Rating

The Centre for Pavement and Transportation Technology at the University of Waterloo and the Ministry of Transportation of Ontario (MTO) have been studying for the past 4 years the suitability of applying automated technologies for network-level evaluations in the province. Three projects have been developed for this purpose. The main results of these studies were a better understanding of available digital technologies, evaluation of the performance of semiautomated and automated technologies, development of new guidelines for pavement distress collection at the network level, design of an adjusted distress manifestation index, and recommendations for the use of semiautomated and automated digital technologies at the network level. The objective of this paper is to present the findings of the third and last phase of the project, Validation and Implementation of MTO Network Level Automated–Semiautomated Pavement Distress Guidelines and Condition Rating Methodology. The scope of the study was to validate and implement in the field MTO network-level distress guidelines and a distress manifestation index for network-level evaluations (DMINL), considering the use of automated technologies. A complete statistical analysis of data collected in the field through manual evaluations and semiautomated and automated technologies is presented. The performance of currently available technologies using network-level distress guidelines was assessed. Finally, from the field validation, distress guidelines were adjusted accordingly and DMINL equations were recalibrated.

Fuzzy set approach to condition assessments of novel sustainable pavements in the Canadian climate

Since the use of pervious concrete pavement structures (PCPSs) is essentially still in the trial stage in Canada, long-term and quantitative pavement condition data are not available. The existing approaches applied to assess pervious concrete pavement structure (PCPS) conditions are ad hoc and suffer from methodological limitations. A fuzzy set technique is proposed herein as an efficient tool for dealing with qualitative and incomplete pavement condition data on distress types, severities, densities, and weighting factors. Using this method, a comprehensive fuzzy condition index was developed based on Ministry of Transportation of Ontario (MTO) methodology and using fuzzy pavement condition data. This fuzzy condition index was converted to a single value that allowed for comparisons of pavement conditions using several ranking techniques. A case study of 24 PCPS sites was utilized to demonstrate how the fuzzy representations of the condition index compared with associated single values. It is shown that this approach can effectively provide extensive condition indices for PCPSs and rank them accordingly, using only limited and imprecise pavement condition data.

The influence of longitudinal cracks on the corrosion protection afforded reinforcing steel in high performance concrete

It is almost impossible to produce crack-free concrete and, therefore, codes of concrete structural design (such as ACI 318 [ACI 318, Building Code Requirements for Structural Concrete, American Concrete Institute: Farmington Hills, MI, USA]) take cracking into account and relate permissible crack widths to exposure conditions. Chloride ingress is significantly enhanced by cracks because the ions can penetrate the concrete cover from the walls of the crack as well as from the outer surface of the concrete [P. P. Win, M. Watanabe, and A. Machida, Penetration profile of chloride ion in cracked reinforced concrete. Cement and Concrete Research, 2004. 34(7): p. 1073–1079]. Thus, while the chlorides reach the steel very rapidly directly through the crack, they also reach adjacent areas of steel more rapidly than in uncracked concrete. The objective of the project was to ascertain whether high performance concrete (HPC) could provide superior protection to Ordinary Portland Cement Concrete (OPCC) even when the structure was cracked parallel to the reinforcing bars (longitudinal cracks) while being exposed to a de-icing salt environment. Two HPCs, as specified by the Ministry of Transportation of Ontario (MTO) [MTO, SSP 904 S13 High Performance Concrete, Amendment to OPSS 904 Construction Specification for Concrete Structures. 1995, Ontario Provincial Standard Specification] were tested: both used blended silica fume cement (Canadian Type 10 E-SF) and had 25% replacement of the cement by either fly ash or ground granulated blast furnace slag. A Class C-2 ordinary Portland cement concrete was used as control. In the case of cracks parallel to the rebar, HPC does not appear to have any beneficial influence on the corrosion of bars, which is not surprising in view of the fact that the whole length of the bar is directly exposed to the environment via the crack. In contrast, HPC has been found to provide better protection for steel exposed to transverse cracks than does OPCC. However, the benefits of HPC are not as great as they are for sound (uncracked) concrete. The reasons for the better protection are (i) the greater resistance of HPC to chloride penetration from the walls of the crack, (ii) its greater tendency to crack healing and (ii) the different crack path in HPC.

The Executive Support System of Ontario, Canada

The Ministry of Transportation of Ontario (MTO), Canada, is currently implementing an asset management business framework (AMBF). The AMBF provides the ministry with an ambitious blueprint for incorporating asset management concepts into its existing business processes. A key component in the AMBF is the ability to integrate results from the ministry's existing management systems. In support of the AMBF, MTO has developed a prototype executive support system (ESS). The ESS is a what-if analysis tool that predicts network performance over time using data from the ministry's pavement and bridge management systems. It enables decision makers to evaluate the relationship between performance and budget and to view results by region, corridor, or functional class. This paper presents the analytical approach used to develop the ESS and describes how it was implemented by MTO. The ESS uses a candidate-based approach to system integration, which enables the integration of any management system capable of generating work candidates and estimating their impact on a defined set of performance measures. The ESS brings together data from these systems and performs an additional level of cross-asset economic optimizations, taking into account user-defined operating assumptions. Although much work has recently been done on the integration of pavement and bridge systems, the objective of this paper is to present a practical example implemented by MTO.

Part 2: Bridge Management Tools: The Executive Support System of Ontario, Canada: Integrating Pavement and Bridge Management Systems

The Ministry of Transportation of Ontario (MTO), Canada, is currently implementing an asset management business framework (AMBF). The AMBF provides the ministry with an ambitious blueprint for incorporating asset management concepts into its existing business processes. A key component in the AMBF is the ability to integrate results from the ministry's existing management systems. In support of the AMBF, MTO has developed a prototype executive support system (ESS). The ESS is a what-if analysis tool that predicts network performance over time using data from the ministry's pavement and bridge management systems. It enables decision makers to evaluate the relationship between performance and budget and to view results by region, corridor, or functional class. This paper presents the analytical approach used to develop the ESS and describes how it was implemented by MTO. The ESS uses a candidate-based approach to system integration, which enables the integration of any management system capable of generatin...

Verification of Network-Level Pavement Roughness Measurements

In 1997 the Ministry of Transportation of Ontario (MTO), Ontario, Canada, switched to international roughness index (IRI) measurements as an indicator of network-level pavement roughness within its pavement management system. Measurement of pavement roughness or ride quality in terms of IRI can be performed with different measuring devices. However, the individual measurements for the same pavement section may vary significantly because of the use of different measuring devices, different longitudinal profiles, and different measuring speeds. Recent evidence obtained by MTO indicates that the longitudinal profile measurements provided by different measuring devices contain systematic differences that range from 0.1 to 1.0 m/km. Such differences in IRIs cause significant concerns for agencies that contract out collection of network-level roughness measurements on a yearly basis. Through the process of verification and comparison of longitudinal profile measurements obtained with different profilers, MTO has gained insight into the functional relationships and factors that affect profile measurements in terms of precision and bias. The verification techniques used to obtain normalized, reproducible, and time-stable IRI measurements for IRIs supplied by different IRI providers are described. Preliminary findings and statistical analyses of IRI values measured on a verification circuit, which is composed of 12 sections with four different pavement types, are provided. In addition, the results of analyses of the various IRI measurements and their impacts on network-level pavement serviceability are discussed.

An approach for auditing highway sections for safety improvements

The objective of this paper is to present a framework for the evaluation of traffic safety improvements on rural highways based on the existing process of the Ministry of Transportation of Ontario (MTO). The framework includes four steps to identify hazardous locations and determine the most feasible improvements. The framework was applied to a 2.2 km segment on Highway 17 (Ontario, Canada) to illustrate how the framework might be implemented in practice. The first step is defining the highway section to be studied. This section is usually a segment that was constructed in the same contract, and its surface conditions require major maintenance. The evaluation of traffic safety on sections that are undergoing major pavement rehabilitation will reduce the overall cost of road maintenance. The second step is to collect and analyze collision data, along with traffic and geometric data. The purpose of this step is to compare the actual number of collisions on the section with the expected long-run mean value and identify the causes of collisions. Based on this analysis, the third step identifies the possible improvements that would eliminate or minimize the number of collisions. These improvements can be as simple as signing improvement or as complex as a major realignment. Finally, these improvements are evaluated economically and environmentally. For this study, several sections of the Ministry of Transportation of Ontario, other ministries and levels of government, and interest groups participated in the evaluation process. A refined MTO process, currently being considered for implementation, is briefly described along with a discussion of its key features.Key words: evaluation, highway improvements, safety, three-dimensional alignments, sight distance, collision frequency.

Sleep Apnea & Automobile Crashes

As a group, patients with obstructive sleep apnea (OSA) are at increased risk of having automobile accidents. Previous studies using actual accident data have used only small numbers of subjects. To determine the rate of automobile accidents in a large population of OSA patients using objective data from the Ministry of Transportation of Ontario (MTO). Retrospective study Academic sleep disorders clinic and laboratory. All cases of OSA polygraphically confirmed between June 1990 and June 1994. Cases of OSA were a priori divided into groups based on apnea-hypopnea index (AHI): (OSA1 - AHI 10-25, OSA2 - AHI 26-40, OSA3 - AHI>40) and driving records were obtained from the MTO. Age and sex matched controls were selected at random from drivers in the MTO driver database who hold passenger vehicle licences. Analysis was restricted to drivers with the same licence class. Primary outcome measure was accidents in the five years preceding diagnosis. Secondary outcome was citations during the same period. There were 155 of 460 OSA patients with one or more accidents compared with 150 of 581 Controls for the same time period (x2=7.7,p<0.01). The rate of accidents/year, for the preceding five years, was 0.07+/-0.14 for Controls versus 0.09+/-0.14 for OSA (p <0.05). This difference could all be accounted for by increased accident rate in OSA patients with the highest AHI (OSA3) (MVA/yr: 0.11+/-0.15, 0.08+/-0.12, 0.06+/-0.14 for OSA groups 3,2,1 respectively) as there was no differences among Control, OSA1 and OSA2 accident rates. OSA patients had twice as many citations as Controls (1.74+/-2.13 vs 0.86+/-1.43 p<0.001) although the types of citation were the same. Increased automobile accidents in OSA may be restricted to cases with more severe apnea (AHI >40). Despite the large sample size (an order of magnitude greater than previous reports using accident data) further study is needed with even larger numbers, including more measures of disease severity and rigorously controlling for driving exposure.