RPUG 2025 Speaker Bios and Abstracts

Abstract:
Improving the Quality of Pavement Surface Distress and Transverse Profile Data Collection and Analysis – Pooled Fund Meeting

This meeting is open to its members and all others interested in this subject.

Abstract:
This is a TPF-5(463) Friction Pooled Fund Meeting to discuss the TPF’s friction studies, including the following presentation:
Physical Interpretation of Contactless Energy-Based Tire-Pavement Friction Models, by Ahmad Alhasan (HNTB)

All TPF members and friends are welcome!

Abstract:
Tire-Pavement friction is a complex phenomenon affected by multiple factors including pavement surface texture, tire geometry and material characteristics, slip ratio, and skew angle. In previous years, the research team presented the framework for energy-based friction models (EBFM). This presentation discusses the physical interpretation of the EBFM model and explores the procedures to derive an approximate closed form solution to estimate the skid numbers (SN) and coefficient of friction (COF). The approach discussed in this presentation will start from the analytical solution for a simple dry rubber block using Persson’s friction model. In this model the surface texture is characterized using wavelets energy as an average representation of the power spectral density (PSD) within a given bandwidth. The physical model parameters will be fine-tuned using the COF measured using the dynamic friction tester (DFT) on 14 pavement sections. Starting from the analytical solution additional functions are added to estimate the impact of different tire configurations and water film in the corresponding SN values adjust the contribution from the Locked Wheel Skid Trailer (LWST). This will allow for defining the contribution of different texture features at different scales corresponding to different pavement material characteristics. The presentation provides insights into the connection between different traction conditions and devices used in the field. Moreover, the model will allow for separating the impacts of test tires, speed, and water film impacts. The EBFM presents a promising approach, given the analytical fundamentals used to derive the prediction models.

Abstract:
This workshop is hands-on and focuses on ProVAL software exercises. It also includes a ProVAL users forum to interact with users for feedback.

Trainers include George Chang(Transtec Group), Dave Merritt, Amanda Gilliland, Subramanian “Subu” Sankaranarayanan

All participants need to bring their laptop computer preinstalled with ProVAL 4.0 and sample files.

This workshop is open to all!

Abstract:

Abstract:

Abstract:

Abstract:
The CT DOT has been collecting and processing our own Pavement Condition Data in house since 2001.
In this presentation I’ll discuss CTDOT department roles and responsibilities and units who participate in this yearly process.
This presentation will focus on our current Quality Control and Quality Acceptance for Pavement Condition Data. What standards and procedures CTDOT have implemented to validate our yearly data.
CT spends a great deal of resources and effort ensuring that our pavement condition data is of acceptable quality.
Topics to be covered for presentation are as follows:
• Discuss CTDOT department roles and shared responsibilities between units who participate in our QC and QA.
• Is our current DQMP fully being executed, what’s working and what needs to be modified?
• CTDOT QC/QA of post-production (processed) pavement condition data.
• In depth dive into automated crack detection within Vision software.
• Is the LCMS crack detection giving us repeatable results?
• Review CTDOT data acceptance reporting and resolution.
This past year CTDOT has been establishing a process to verify the efficiency of automated cack detection. This process involves using the pavement tool within the software and making a visual assessment of all cracking parameters. I will discuss the findings we’ve discovered after reviewing a sample set of processed crack maps. How we summarized our findings and incorporated this process into our current QC/QA procedures.

Abstract:
Advanced computing and 3D sensing technology have enabled transportation agencies to utilize 3D pavement data to extract cross-slope information, which is critical for roadway safety. However, there is a need to critically evaluate the accuracy of cross-slope measurements derived from 3D pavement data. This study develops a methodology to quantitatively assess cross-slope measurement accuracy by comparing it with ground reference values obtained using the manual straight-edge method. The proposed methodology includes: 1) gathering 3D pavement surface test data and ground reference data, 2) preparing and processing 3D pavement data in LAS format, 3) registering and aligning test and ground reference measurements to ensure spatial consistency, 4) comparing the datasets to evaluate the cross-slope measurement accuracy, and 5) reporting accuracy and interpreting the results. Data collected from a Texas test site in both standard file format (PSI) and LiDAR file format (LAS) have been used. Preliminary results indicate that cross-slope measurements derived from LAS files achieve an R² value of 0.95 when compared to manually measured cross-slope values. This study demonstrates the feasibility of using 3D laser technology for accurate cross-slope measurement and validates the proposed methodology for transportation agencies to assess cross-slope measurement accuracy.

Abstract:

Abstract:
Crack detection and classification methods have focused on 2D and 3D images as the starting point for analysis. However, different sensors and imaging systems measure different raw data, complicating analysis and research efforts. It is proposed that a crack map is a better starting point for most applications including for AI-based classification methods. A crack map represents an intrinsic, physical property of the pavement, can be obtained from many sensors, and can be processed to meet any agency’s classification requirements. This presentation proposes an international standard crack map representation, adapting an existing standard defined by the Open Geospatial Consortium for geometry objects. This crack map representation is being proposed as a new ASTM standard with the goal of promoting further innovation in cracking analysis.

Abstract:
The presentation will cover the results of a master thesis that investigated the potential for leveraging connected vehicle data for enhanced road condition monitoring in the state of Virginia. Using a manufactured-provided advanced data collection technology, which gathers road surface information from standard connected passenger vehicles, researchers compared this modern approach with legacy data collection methods, such as dedicated road profiling vehicles. A sample of data from the a district was used to compare the connected-vehicle estimated values to the standard International Roughness Index (IRI) values used by VDOT. Data collected using both methods were matched to a common linear referencing system. Subjective visual comparison showed that both data sets had similar trends, highlighting roads with rough sections. A quantitative analysis performed to compare the average results of the two methods on a sample of uniform sections, showed a high correlation.
The results confirmed that CV technology can be a valuable tool to complement traditional ride quality data collection methods and can be used for pavement evaluation purposes. Although estimated CV roughness values were slightly higher than the reference IRI values, poor road surfaces can still be located using the estimated CV-estimated roughness data. Additionally, since CV data is being collected all year long, agencies have access to the latest data at any time, without needing to deploy specific employees and specialized equipment to collect the data, making it a more convenient method. The data can be used for predictive maintenance planning, validation of performed work, and an overall road condition ranking. This collaboration highlights the potential for scaling connected vehicle-based data collection across state and regional networks, reducing reliance on infrequent, and resource-intensive legacy methods while unlocking new efficiencies and applications.

Abstract:
Transportation agencies increasingly rely on automated cracking distress detection systems due to advancements in pavement imaging technologies and computational methods. However, the accuracy of extracted cracking data is often compromised by factors such as sensor degradation, vehicle vibration, and algorithmic errors. Currently, no standardized methodology exists to thoroughly evaluate data quality from these systems, limiting agencies’ ability to ensure reliability of pavement management. This study addresses this gap by proposing a two-tiered methodology to assess data quality and pinpoint error sources: 1) End-to-end evaluation: This approach establishes ground reference (GR) cracks through manual crack mapping on paper sheets and camera-captured crack augmentation on pavement surfaces, enabling direct comparison between detected cracks and real-world conditions; 2) Decomposed analysis: This method isolates errors into sensor-related (e.g., imaging quality) versus algorithmic (e.g., detection accuracy) issues using machine-fabricated GR boards and manual cracking annotations of pavement images. A case study was conducted in collaboration with the Florida Department of Transportation (FDOT) to explore and validate the proposed methodology. The results of this study will establish a standard and practical cracking data quality assessment methodology for transportation agencies to implement during routine data quality assurance checks.

Abstract:
3D sensors (LCMS) have been available and used by the road inspection community for over 15 years now to acquire 1mm resolution 3D images of road surfaces at highway speeds. The first use case for this data was to automate the detection and analysis of cracks on HMA type asphalt surfaces. Early crack detection algorithms were based on traditional machine vision algorithms using rule-based algorithms and statistical analysis of the data to make decisions as to the presence of defects.

With the advent of AI tools and processing capabilities in the last 5 years it is interesting to for us to evaluate how this technology can be used to possibly improve the performance of traditional machine vision. With this goal in mind, we started experimenting with AI for this purpose and this paper presents some of the more interesting results that were found and lessons learned such as:

When is it best to use AI
When is it not appropriate to use AI
The AI solution development and training process

Results:

• AI is really good at detecting and classifying objects or defects but less good at making measurements;
• IRI, macro-texture, rutting, faulting, raveling, road geometry for example are not good candidates for AI.
• AI for crack detection shows both issues and potential.
• AI was successfully shown to improve the detection of sealed cracks, joints, man-made objects and pumping.
• Images presenting both traditional and AI processed results will be side by side.

Abstract:

Abstract:

Abstract:
The Illinois Department of Transportation (IDOT) and the Illinois Center for Transportation (ICT) conducted a longitudinal road profiler comparison experiment at the Illinois Certification and Research Track (ICART) October 28 – November 16, 2024. The research sought to refine, demonstrate, and document procedures for regional profiler certification.
The experiment included accuracy and repeatability testing of conventional high-speed and lightweight inertial profilers, novel road profilers, walking-speed profilers, and high-speed profilers with special systems for working at low speed, during braking, and through stops. The Federal Highway Administration (FHWA) Benchmark Profiler (BP) and the FHWA Urban and Low-Speed Profiler (ULSP) collected reference measurements. The testing took place on six pavement surface types at ICART of diverse material and surface texture type.
Forty-two profile measurement devices participated in the experiment. This presentation: (1) describes the testing and analysis methods used in the experiment, (2) discusses the results for high-speed and lightweight inertial profilers, novel road profilers, and walking-speed profilers, (3) provides observations of results for runs from high-speed inertial profilers conducted at different speeds, and (4) demonstrates the sensitivity of the results to height sensor footprint type and the algorithm used to establish a single height value from each reading from a line laser (i.e., the bridging filter).
Note: Results for testing at low speed, during braking, and through stops are offered in a separate presentation. If the community is sufficiently interested, the footprint issue could be covered in more detail in a separate presentation also.

Abstract:
This presentation discusses testing performed at the Illinois Certification and Research Track (ICART) to establish the accuracy of longitudinal road profilers at low speed, during braking, and through stops.
Recent additions to AASHTO R-56 describe testing and analysis methods for establishing the accuracy of profilers during operational conditions that typically introduce errors into the measurements of conventional inertial profilers. The methods were incorporated into AASHTO R-56 for two reasons. First, the new methods provide a way to certify profilers with advanced technologies for collecting valid profile at low speed, during braking, and through stops. Second, the new methods provide a way to determine the sensitivity of conventional profilers to these conditions, and set thresholds for minimum travel speed, maximum braking, and the range of invalid profile in the area near a stop.
A profiler comparison study was conducted at ICART in the fall of 2024. The study included pilot certification testing of eight profilers using the new methods from AASHTO R-56. The testing included four types of commercial profilers with advanced technology intended to measure valid profile at low speed and through stops. The testing also included three conventional inertial profilers.
This presentation provides technical background the way low speed, braking, and stops introduce errors into measurements from conventional inertial profilers. Examples will be provided from the testing of conventional inertial profilers. The presentation will review the testing and analysis methods in AASHTO R-56, and provide practical advice for their implementation. Finally, the presentation will report on the status of commercial profilers designed to operate at low speed and through stops.

Abstract:
The Illinois Department of Transportation implemented it first version of the International Roughness Index for construction acceptance in 2021. This specification greatly increased the number of treatments and facilities that are required to be tested. To help implement the specification, the Bureau of Research developed several testing guidelines to ensure similar testing procedures between contractors and the Department. Two of the test procedures detail the analysis that Research staff will use to process all the data, while the other two test procedures detail the requirements for the actual testing.
Since 2021, IDOT has collected and analyzed data on over 400 contracts. This paper details the lessons learned over the past four years and insights that can be gathered from the data and analysis. This paper will cover IDOT’s current specifications, test procedures, and analysis.

Abstract:
CDOT’s pavement smoothness specifications had been unchanged for several years and required updating to reflect current best practices and drive industry improvements. Analysis of recent initial smoothness results highlighted the need for updated standards, prompting CDOT to undertake a collaborative revision process.
This presentation will delve into the key changes implemented in the updated specification. It will detail the specific metrics and acceptance criteria that have been revised to ensure higher quality pavements. Furthermore, the presentation will explore the rationale behind these changes and the long term thought process.
A crucial aspect of the specification’s development was the close collaboration between CDOT and both the asphalt and concrete industries. This presentation will highlight the collaborative approach taken, emphasizing the valuable input provided by industry stakeholders. It will showcase how this partnership ensured that the new specification is not only technically sound but also practical and achievable within the current construction landscape. The discussion will focus on how the collaborative process fostered mutual understanding, addressed industry concerns, and ultimately resulted in a specification that serves the best interests of both CDOT and the traveling public.

Abstract:

Abstract:
Real-Time Smoothness (RTS) technology is a tool for helping to improve the as-constructed smoothness of concrete pavements. RTS technology measures the profile of a concrete pavement slab behind the paver, providing real-time feedback on smoothness during the actual paving operation. This allows contractors to diagnose potential issues in the paving operation that affect smoothness in order to make equipment/process changes immediately, rather than waiting for feedback after the hardened pavement is profiled. From 2014-2024 FHWA provided an opportunity for agencies and concrete paving contractors to evaluate commercially available RTS systems under an equipment loan program facilitated by the National Concrete Pavement Technology Center at Iowa State University. Under this equipment loan program RTS technology was demonstrated on 25 projects in 18 states. These evaluations provided tremendous insight into the benefits, limitations, and overall usefulness of RTS technology and resulted in a number of contractors purchasing their own RTS system following the equipment loan. This presentation will summarize the current state-of-the-practice for RTS for concrete paving and some of the key lessons learned from the equipment loan program. In addition, a value proposition highlighting the value of RTS to contractors, agencies, and the traveling public with respect to cost versus benefits will be presented.

Abstract:
The presentation would be relevant and of interest to users in the following topic areas:
• Profile Data Collection Standards and Certification
• Network Pavement Surface Measurements
• Pavement Roughness and Ride Comfort
• Pavement Surface Characteristics Data Quality Management
The single laser inertial profiler, also known as the “High Speed Profiler (HSP),”, has delivered good quality profile data for decades. However, its use is generally limited to freeways and highways. Its utility is limited in urban or suburban areas since it can’t tolerate speed variation, low speeds or stopping that would be expected in traffic congestion and at traffic controls in and around cities. By replacing the single laser and vertical accelerometer with dual-lasers and inclinometer we can have a profiler that is totally independent of speed and acceleration and has unlimited use cases. This represents the first paradigm shift in profiling technology in the last 70 years.

The presentation would compare HSP with dual-laser methods and reflect on the history of development of dual laser technology from its earlier use in different reference profilers up to recent highway speed developments, particularly the measurement of vehicle pitch using various inertial methods. We would explain how the technology works and the benefits of the dual-laser direct measurement of profile slope rather than elevation, the obvious one being the total independence from speed and acceleration. Measuring profile slope directly, rather than elevation, may at first appear counterintuitive but it makes a lot of sense since surface profile roughness analysis actually uses slope rather than elevation. We would discuss an integrated dual technology solution that computes both HS and dual-laser profiles for constant speed runs from the same raw data and cross-correlates them internally to provide greater user performance confidence. Direct comparisons of HSP and dual-laser profiles for the same constant speed run will be presented. We would also show a wide range of stop and go profiling cases supported by velocity and acceleration for braking and stopping, including runs collected at the ICART test track. The current technology and software enables smooth, accurate profile collection that is glitch-free regardless of hard braking, long durations stops or hard acceleration.

Abstract:
Lidar scanning technology is increasingly popular for roadway construction. This presentation will provide a review of the scanning process of the existing surface, real-time data collection during the milling and paving operations, and final surface reality. It will explore options of what can be done with lidar data and review the proven results. When leveraging scanning technologies field crews come home safely and projects move forward with more data than ever before. It would also present a case study from a contractor’s perspective with real-world project data and the lessons learned to improve construction smoothne4ss quality.

Abstract:

Abstract:
Classifying slab states based on cracking patterns, such as longitudinal, transverse, corner cracking, or a combination of these (i.e., shattered), is critical for determining the cost-effective maintenance strategy for Jointed Plain Concrete Pavements (JPCP). With the availability of crack maps derived from high-resolution 3D pavement surface images, there is a significant opportunity to automate the classification of JPCP slab states based on cracking. While previous studies have demonstrated the feasibility of deep learning (DL) for slab classification, the interpretability and practical implementation of complex DL models remain challenging. This study proposes a novel methodology for automated and interpretable JPCP slab classification, aligned with a revised LTPP distress definition. An innovative Crack Vector Model (CVM) and a multi-dimensional cracking extent projection method are introduced to extract geometric features from crack maps, enabling the use of tree-based models for effective slab classification. The results show a weighted F1-score of 86.12%, comparable to that of a DL model, while tree-based models offer greater interpretability and ease of implementation. The improved interpretability and simplicity of this method facilitates its practical application in slab-based JPCP classification, supporting transportation agencies in achieving cost-effective, precision pavement maintenance.

Abstract:

Abstract:
Each of the 3 state transport agencies situated along the east coast of Australia utilizes sideways force coefficient measurement technology to measure the skid resistance of their state highway and major road networks. At present there are 3 of these devices based in Australia, 2 of which have been operated by transport agencies for more than 40 years and the other operated by NTRO since 2018.

However, until 2023 there had never been an opportunity to bring the 3 devices together to undertake a comparison trial to determine the consistency of their outputs. This had been assumed but never proven due to each device using the same technology licensed by TRL. Several test locations consisting of different surface types (concrete, asphalt and spray seal) were selected in southern New South Wales to perform the testing.

This presentation details the history of using sideways force measurement technology in Australia along with the outcomes of the correlation trial. Also discussed will be some of the issues faced in performing the trial and plans for skid resistance measurement in Australia in the future along with some of the policies affecting skid resistance measurement.

Abstract:
Each of the 3 state transport agencies situated along the east coast of Australia utilizes sideways force coefficient measurement technology to measure the skid resistance of their state highway and major road networks. At present there are 3 of these devices based in Australia, 2 of which have been operated by transport agencies for more than 40 years and the other operated by NTRO since 2018.

However, until 2023 there had never been an opportunity to bring the 3 devices together to undertake a comparison trial to determine the consistency of their outputs. This had been assumed but never proven due to each device using the same technology licensed by TRL. Several test locations consisting of different surface types (concrete, asphalt and spray seal) were selected in southern New South Wales to perform the testing.

This presentation details the history of using sideways force measurement technology in Australia along with the outcomes of the correlation trial. Also discussed will be some of the issues faced in performing the trial and plans for skid resistance measurement in Australia in the future along with some of the policies affecting skid resistance measurement.

Abstract:
High friction surface treatments (HFST) were first developed during the 1950s and have evolved into a combination of epoxy resign and calcined bauxite that is generally placed along rural curves to prevent roadway departure and wet weather crashes. The Kentucky Transportation Cabinet (KYTC) Highway Safety Improvement Program (HSIP) has been utilizing a data-driven HFST program since 2009 to prevent wet weather and roadway departure crashes. A common procedure, that KYTC started their HFST program with, is screening curves statewide using wet weather crash data. The potential problem with that method is that it assumes HFST is only needed along curves. KYTC started collecting continuous pavement friction data in 2020, which enables a proactive safety performance function (SPF) model to predict the safety benefit of HFST along all interstate, parkway, state primary, and state secondary routes. This presentation will walk through how KYTC identifies candidate locations for HFST and how the program has evolved throughout its history. The most recent identification process utilizes a mixture of pavement friction data, pavement resurfacing lists, crash data, a proactive safety performance function (SPF) model, and field visits. A proactive SPF model has been the data-driven motivator advising KYTC to place HFST near intersection. In 2009, KYTC started placing HFST along rural curves where wet weather crashes were occurring frequently. In 2014, KYTC started placing HFST along intestate ramp curves and rural curves where wet weather crashes were occurring frequently. In 2015, KYTC awarded its first HFST project near an intersection. More recently, 75% of the 28 HFST projects that KYTC awarded during 2024 were near an intersection.

Abstract:
The Mississippi Department of Transportation (MDOT) in cooperation with ARRB Systems has completed the first network level friction collection in the United States using the iSAVE device. The iSAVE is a self-contained friction testing device which measures sideways friction force continuously via a specific narrow wheel set at an angle to the direction of travel. There are several advantages iSAVE has in comparison to traditional lock wheel friction testing devices. The presentation will serve to introduce the audience to the iSAVE, to discuss the advantages of this device, to showcase how data was leveraged to identify and remedy safety concerns within the State Highway System (SHS), and to share lessons learned from this collection process.
A specific case study will be presented where the crash rate of an interstate system interchange ramp suddenly and dramatically increased. The iSAVE was instrumental in collecting friction data at this location because the curvature of the ramp precluded traditional lock-wheel testing. As part of this analysis a procedure was developed to translate the recorded friction to the running speed of the ramp and compare the measured friction to the AASHTO Green Book recommended friction design value specified for curvature and superelevation. This resulted in identifying a location where the superelevation had been unintentionally reduced during a mill and inlay project.
MDOT also completed an International Road Assessment Progamme (iRAP) for a portion of the SHS. The iRAP assessment was conducted using the iSAVE data, video logs, and other data on approximately 104-miles of the surveyed SHS. The iRAP assessment assigns a 1-5 star safety rating to each road and estimates the socioeconomic costs associated with crashes. The analysis also provides recommendations for improving safety and estimates for lives saved for any given treatment.

Abstract:
Continuous tire-pavement friction measurement (CPFM) is crucial to the management and sustained success of High Friction Surface Treatments (HFST). HFST have been identified by FHWA as “pavement treatments that dramatically and immediately reduce crashes, injuries, and fatalities associated with friction demand issues…”. The success of HFST, however, is not guaranteed; HFST can fail for many reasons, including but not limited to placement over pavement in poor or questionable condition (cracking, rutting, raveling, etc.), inadequate material mixing or quality, and installation errors. HFST can fail prematurely in its entirety, or in patches, resulting in low-friction areas. The use of continuous tire-pavement friction measurement equipment can be used to identify the boundaries of HFST, as well as play a vital role in assessing the entirety of the treatment’s effectiveness. Regular or systematic CPFM helps monitor potential issues by confirming the treatment’s performance, facilitating early detection of degradation, and aiding in maintenance planning by identifying when intervention is necessary. This detailed friction data supports data-driven decision-making within a pavement friction management program, and helps transportation agencies maintain safe and effective roadways.

Abstract:
After the introduction of Texture Signature Performance Index (TSPI) in 2024, further developments have been made over the last year to integrate this pavement performance evaluation tool into pavement management groups’ condition assessment programs. Since introducing TSPI, several pavement management groups/agencies have provided “real world” data and user feedback to help steer the development of the Texture Signature Performance Index as it relates to the decision-making process for pavement condition evaluations, as well as safety evaluations. This presentation will review the “real world” TSPI data and how it is being evaluated to make pavement management/maintenance decisions, as well as how and where pavement friction tests should be performed.

Abstract:
After the introduction of Texture Signature Performance Index (TSPI) in 2024, further developments have been made over the last year to integrate this pavement performance evaluation tool into pavement management groups’ condition assessment programs. Since introducing TSPI, several pavement management groups/agencies have provided “real world” data and user feedback to help steer the development of the Texture Signature Performance Index as it relates to the decision-making process for pavement condition evaluations, as well as safety evaluations. This presentation will review the “real world” TSPI data and how it is being evaluated to make pavement management/maintenance decisions, as well as how and where pavement friction tests should be performed.

Abstract:
Lane departure crash patterns can be identified in various ways with one of the methodologies being an increased frequency of lane departure crashes and/or crash rates during wet road conditions. Lane departure crashes and crash rates may increase when the surface is wet because skid resistance can be affected and reduced under these conditions. Research shows that newly overlaid pavements could negatively impact friction and macrotexture, increasing crash risks. This study used field and laboratory experiments alongside statistical analysis to: (1) develop friction and macrotexture performance models for North Carolina roadways, (2) propose friction and macrotexture performance thresholds, and (3) identify asphalt mixture factors affecting macrotexture and friction. The models accurately predicted in-service friction and texture, validated with independent measurements. Thresholds were determined by analyzing crash rates in relation to available friction and texture. Recommended thresholds include:
• Macrotexture: Investigatory level 0.80 mm, Intervention level 0.60 mm (AMES AccuTexture 100).
• Friction: Investigatory level 0.57 for non-interchanges and 0.65 for interchanges. Intervention level: 0.43 for non-interchanges, 0.49 interchanges (Moventor BV-11 at 60 mph).
These findings apply mainly to high-speed, controlled-access roadways. Data from North Carolina’s highway network, combined with threshold values and model predictions, were used to assess maintenance and rehabilitation strategies. Regardless of the maintenance scenario or interest rate used, all benefit-cost analyses yielded a ratio greater than one, indicating cost-effective improvements in friction and texture management.

Abstract:
The condition of pavement texture properties is crucial in assessing pavement safety. The NHTSA estimated that the fatality rate for the first three months of 2024 declined to 1.13 fatalities per 100 million vehicle miles traveled, compared to 2023. While serious crashes frequently occur due to driver’s behavior, the contribution of pavement surface properties cannot be disregarded. Highway safety is impacted by pavement surface characteristics, especially friction and macrotexture. Yet, maintaining adequate friction is essential for crash prevention, particularly on higher speed traffic. As speed varies, tire-pavement interaction changes, making the relationship between macrotexture and friction increasingly complex. While evaluating friction and macrotexture as individual variables is inadequate to comprehensively capture the interaction of tire-pavement surfaces, implementing this interaction according to the speed limits and facility type should be considered.
A novel approach is proposed to investigate the speed-dependent frictional behavior of pavement surfaces using network level Continuous Friction Measurements (CFMs). This incorporates the combined effects of macrotexture and low slip-speed friction, to assess Safety Performance Functions (SPFs). The methodology involves determining friction values (FRS) in accordance with the ASTM E1960-07. The FRS values are refined using macrotexture-based speed constant, adjusting across a range of operating speeds (i.e., 40, 55, and 70 mph) according to the roadway’s functional classification. From the analysis, comparing the behavior of various asphalt surface mix types under the influence of the combined index with individual contributions, revealed that the estimated available friction at different speeds significantly combines the effect of friction and macrotexture on freeway and primary crashes. Findings enabled the development of Crash Modification Factors (CMFs), for estimation of crash risk potential due to changes in both friction and MPD. Further, analysis confirms how certain mix types demonstrate heightened sensitivity to combined effects, which has critical implications for mix design and surface treatment strategies.

Abstract:
Macro texture is an important property of highway surfaces as it contributes to various characteristics such as road noise and high-speed friction. This paper is primarily concerned with the influence of macro texture on high-speed friction. Macro texture is typically measured using lasers and characterized within a certain wavelength band. The measured texture profile is processed using an algorithm specified in standards to produce an MPD value. With regards to high-speed friction, the significance of macro texture is the ability of the surface to disperse water from the tire/road contact patch area. Consequently, an alternative measure to MPD was developed which considers the area between the tire and surface rather than certain “peak” values as specified in the MPD algorithm. This paper describes this new measure and presents the results of the comparison between it and MPD.

Abstract:

Abstract:
Macrotexture is a critical pavement surface property that significantly influences pavement safety. Recently, the Federal Highway Administration (FHWA) developed a draft laboratory test protocol for measuring the macrotexture of dense-graded asphalt mixtures using a laser texture scanner. To establish this protocol as a standard test method, it is essential to evaluate its precision, including both repeatability (variability across replicates) and reproducibility (variability across different laboratories). To this end, an interlaboratory study, involving five laboratories, is conducted. A set of specimens, covering a wide range of mean profile depth (MPD) values (0.2mm < MPD < 1.2mm), was identified. Then, these specimens are tested by the five laboratories with multiple replicates to estimate the variability associated with this test protocol. The results provide a precision estimation for the proposed test method, supporting its standardization as a reliable laboratory macrotexture testing procedure. The interlaboratory study generated a large database of asphalt mixture macrotexture. Based on this dataset, a machine learning model incorporating computer vision technologies is developed to enhance macrotexture characterization. The model effectively extracts critical features of macrotexture and creates a low-dimensional representation, which provides a novel approach for macrotexture characterization. The results demonstrate that the proposed method offers a promising assessment of macrotexture compared to traditional methods, such as mean profile depth (MPD), with the goal of better understanding of pavement surface characteristics.

Abstract:
Macrotexture is a critical pavement surface property that significantly influences pavement safety. Recently, the Federal Highway Administration (FHWA) developed a draft laboratory test protocol for measuring the macrotexture of dense-graded asphalt mixtures using a laser texture scanner. To establish this protocol as a standard test method, it is essential to evaluate its precision, including both repeatability (variability across replicates) and reproducibility (variability across different laboratories). To this end, an interlaboratory study, involving five laboratories, is conducted. A set of specimens, covering a wide range of mean profile depth (MPD) values (0.2mm < MPD < 1.2mm), was identified. Then, these specimens are tested by the five laboratories with multiple replicates to estimate the variability associated with this test protocol. The results provide a precision estimation for the proposed test method, supporting its standardization as a reliable laboratory macrotexture testing procedure. The interlaboratory study generated a large database of asphalt mixture macrotexture. Based on this dataset, a machine learning model incorporating computer vision technologies is developed to enhance macrotexture characterization. The model effectively extracts critical features of macrotexture and creates a low-dimensional representation, which provides a novel approach for macrotexture characterization. The results demonstrate that the proposed method offers a promising assessment of macrotexture compared to traditional methods, such as mean profile depth (MPD), with the goal of better understanding of pavement surface characteristics.

Abstract:
The Road Profile Users Group meets annually to discuss the tools and processes for addressing highway safety and ride quality issues. Diamond grinding and safety grooving have been proven time and time again to be able to remediate these issues in a cost-effective and practical way. Since diamond grinding is a process, and not a material, it is the only highway treatment that has, in some instances, proven to be cost and carbon negative. This has been quantified using the IGGA’s cost and carbon calculator built in partnership with the MIT Concrete Sustainability Hub. As good stewards of the public domain it is important for owner agency engineers to consider the economic and environmental impact of the processes they chose to deploy. The ability to choose a fast, low-cost, sustainable remediation tactic for ride quality issues that results in real savings on fuel resources for highway users is an important tool for RPUG attendees to be informed of.

Data from various sources has also shown that safety grooving can substantially improve friction numbers of pavement, even in instances when inadequate friction stone was used. PennDOT collected data on soft limestone concrete pavement where they experienced rapid polishing of diamond ground texture. Areas where safety grooving was included resulted in nearly double the skid number after 6 years of service. Reduction in hydroplaning and directional stability also contribute to the significant improvement of safety on roadways. As safety is one of the highest concerns of highway engineers simply collecting skid numbers is not a sufficient effort if there is not a quick and cost-effective response to address the immediate risk to the traveling public. Safety grooving is inexpensive and can be deployed quickly on all pavement surfaces to address immediate friction concerns.

I feel as though both topics are critical for highway engineers to be aware of as diamond grinding and safety grooving can remediate virtually all of the distress types discussed at RPUG conferences.

Abstract:
Pavements condition data is a critical input to agencies decision in managing and maintaining pavement assets and improving the design practices for the future. Despite the long history of pavement condition data collection and evaluation, there are still some uncertainties and challenges in handling the variability in the collected data over time to build performance models and the impact of changes to the data collection technologies and standards on the performance models. In this presentation, we will present a comprehensive review of the sources of variability in pavement condition data and the impact they have on performance projections uncertainties and project selection and decision making. The review will be illustrated using statewide condition data records, some going back 50 years, including ride quality, rutting, and Condition Rating Survey (CRS) collected for the network in Illinois. The analysis will illustrate the quantitative impacts of the data variability and the changes in data collection procedures on network performance models and decision-making process. In the second half of the presentation, we will discuss the current advancements and challenges with using machine learning and artificial intelligence (ML/AI) in pavement condition evaluation and performance modeling. The presentation will focus on what ML/AI can and cannot provide in its simple form and challenges leading to misuse and misapplication of ML/AI as “black box” tools. A solution will be introduced to performance and engineering models using Bayesian Networks as a pseudo physics-guided ML model. The example will address the workflow in using these algorithms and procedures to debug such application and avoid unrealistic models that do not conform with engineering behavior and physical reality

Abstract:
Jointed Plain Concrete Pavements (JPCPs) are predominantly constructed on high-traffic volume routes such as interstate highways and major freight corridors, where timely and effective maintenance and rehabilitation (M&R) planning is critical. Traditional M&R project planning and design methods rely on site inspections, low-speed windshield surveys, and limited field measurements. These approaches pose significant safety risks for agency personnel and motorists, particularly when assessing inner lanes of multi-lane highways with heavy traffic, while also being time-consuming, tedious, and prone to inaccuracies in identifying all required repairs. Additionally, generating optimal design plans is hindered by the lack of critical information, such as joint spacing and previous slab replacement history, along with the need to adhere to multiple design constraints. With the widespread availability of 3D pavement surface data through agency-owned systems or on-demand vendor services, there is a significant opportunity to enhance and automate the M&R planning and design process for JPCPs.
This presentation introduces a novel methodology that leverages 3D pavement surface data to automate JPCP M&R design plans and quantity estimation. The approach involves developing a comprehensive slab-based inventory using automated joint detection results, integrating automatically detected crack maps with classified slab cracking states, and extracting joint faulting and spalling conditions. These insights enable precise identification of required M&R actions at the slab and joint level, including full or partial slab replacement, joint repair, crack sealing, and diamond grinding. The system generates spatially referenced repair design plans, optimized slab replacement joint layouts, accurate quantity estimates, and supporting information for traffic control planning. A case study on a project along Interstate 16 (I-16) from MP 22 to MP 12 near Macon, GA, in the westbound direction, will be presented to showcase the developed methodology using 3D pavement surface data obtained by the Georgia Tech Sensing Van. By extending the use of 3D pavement data beyond network-level condition assessment to project-level applications, this methodology enhances decision-making, improves safety, and streamlines M&R project execution for transportation agencies.

Abstract:
Traditional road monitoring methods rely on dedicated survey vehicles to assess pavement conditions, often requiring extensive resources and limiting coverage in both frequency and number of lanes. A network-wide screening approach using connected vehicle data offers a scalable, cost-effective alternative, enabling continuous road condition assessment across entire networks and the identification of schemes for deploying specialized machinery.
By leveraging data from vehicles equipped with standard sensors—such as accelerometers, wheel speed sensors, and GPS—real-time road roughness, friction, surface anomalies, and low structural bearing capacity can be detected. This distributed sensing approach provides broad, high-frequency coverage, allowing transportation agencies to identify areas requiring project level investigation. It enables proactive maintenance planning by highlighting sections with significant changes while avoiding unnecessary assessments of well-maintained roads. By integrating connected vehicle data into asset management workflows, agencies can allocate resources more efficiently.
Early trials of network-wide screening demonstrate its potential to reduce costs, improve decision-making, and enhance road safety by detecting hazardous conditions before they escalate. As adoption increases, connected vehicle data will play a crucial role in transforming road management from a reactive to a data-driven, predictive approach, ensuring smarter, more sustainable infrastructure management.

Abstract:
Traditional road monitoring methods rely on dedicated survey vehicles to assess pavement conditions, often requiring extensive resources and limiting coverage in both frequency and number of lanes. A network-wide screening approach using connected vehicle data offers a scalable, cost-effective alternative, enabling continuous road condition assessment across entire networks and the identification of schemes for deploying specialized machinery.
By leveraging data from vehicles equipped with standard sensors—such as accelerometers, wheel speed sensors, and GPS—real-time road roughness, friction, surface anomalies, and low structural bearing capacity can be detected. This distributed sensing approach provides broad, high-frequency coverage, allowing transportation agencies to identify areas requiring project level investigation. It enables proactive maintenance planning by highlighting sections with significant changes while avoiding unnecessary assessments of well-maintained roads. By integrating connected vehicle data into asset management workflows, agencies can allocate resources more efficiently.
Early trials of network-wide screening demonstrate its potential to reduce costs, improve decision-making, and enhance road safety by detecting hazardous conditions before they escalate. As adoption increases, connected vehicle data will play a crucial role in transforming road management from a reactive to a data-driven, predictive approach, ensuring smarter, more sustainable infrastructure management.

Abstract:
The Texas Department of Transportation (TxDOT) has embarked on an innovative project, titled “”Artificial Intelligence for Pavement Condition Assessment from 2D/3D Surface Images,”” to enhance pavement condition evaluation through the application of Artificial Intelligence (AI) and Machine Learning (ML) technologies. This effort represents one of the earliest statewide explorations into leveraging AI/ML for pavement distress detection.
This presentation outlines the methodologies employed to develop and validate a pavement distress detection model at the state level. Key steps include: 1) building an extensive pavement image library, 2) developing AI/ML models for various distress types, and 3) validating the models’ performance in both controlled testing and real-world scenarios. Notable challenges encountered throughout the project are addressed, such as managing data quality and quantity, ensuring model scalability, and establishing robust evaluation methods. Findings indicate that data imbalances, diverse pavement conditions, image quality, and annotation quality are critical factors influencing detection accuracy. Results from controlled test datasets indicate detection accuracies exceeding 80% for well-represented distress types, highlighting the potential of AI/ML for large-scale deployment. However, performance declines for less common distress types in Joint Concrete Pavement (JCP) and Continuously Reinforced Concrete Pavement (CRCP), including punchouts and corner breaks. Besides, real-world case testing reveals variations in performance, underscoring the need for continued refinement of both data collection practices and modeling techniques. The detection results over different pavement types with different methods are also presented, with corresponding analysis. This presentation shares key insights and recommendations drawn from these results, offering valuable guidance for practitioners looking to adopt AI/ML solutions for large-scale pavement condition assessment.

Abstract:
Pavement surface macrotexture significantly influences roadway safety and performance by affecting the interaction between vehicle tires and the pavement surface. Macrotexture refers to the large-scale texture of the pavement surface, which is a critical factor that affects pavement surface conditions. This research investigates the use of 3D laser technology to evaluate macrotexture of pavements. In this study, a Laser Crack Measurement System (LCMS) was employed to collect 3D pavement data at posted speeds. The Mean Texture Depth (MTD) values were calculated using the manufacturer’s software. The scanned pavement areas were divided into five transverse zones per lane: outside left wheel path, left wheel path, between wheel paths, right wheel path, and outside right wheel path. Longitudinally, the areas were segmented into 1-foot intervals, resulting in rectangular segments for analysis. The MTD value was calculated for each segment and georeferenced to the roadway’s milepost and GNSS location. The MTD results revealed several locations on the tested road with low values. Further investigation of these areas, using high-resolution imagery, identified pavement bleeding as the underlying distress. Additionally, skid testing results for the tested road, performed by others, were reviewed. The skid numbers were calculated in accordance with ASTM E-274 and ASTM E-524. A correlation analysis between the MTD values and skid numbers demonstrated a strong relationship, particularly in areas with both low skid numbers and low MTD values. This research highlights the effectiveness of 3D laser technology in identifying pavement macrotexture deficiencies and its correlation with skid resistance, providing valuable insights for roadway safety assessments.

Abstract:

Abstract:

Abstract:

Abstract:
TRB Standing Committee on Pavement Surface Properties & Vehicle Interaction mid-year meeting

This meeting is for its members and all others interested in this subject.

Abstract:
This is TPF-5(537) Profile Pooled Fund Meeting, including the following presentations:
• Testing with the Benchmark Profiler and the Urban and Low-Speed Profiler at ICART, by Steve Karamihas (APTech)
• Contractor Pavement and Bridge Profiling Verification, by Max G. Grogg (APTech)

This meeting is for its members and all others interested in this subject.

Abstract:
This presentation discusses upgrades to the Benchmark Profiler (BP) and the Urban and Low-Speed Profiler (ULSP) and testing with them at the Illinois Certification and Research Track (ICART).
The BP was developed as a basis for establishing the accuracy of reference devices for measuring longitudinal profile, and had been used for that purpose in five experiments in 2009 through 2015. The ULSP was developed in 2015 to demonstrate a method for valid profile measurement at low speed, during braking, and through stops. The BP and ULSP have been upgraded and tested in an effort to demonstrate their readiness as interchangeable reference profile measurement devices within the profiler certification program at ICART. The BP will provide “benchmark” profiles at least once each certification season, and the ULSP will provide reference quality profiles on regular intervals and frequently on jointed concrete pavements that exhibit changes throughout the day.
This presentation describes the measurement principle of the BP and ULSP, the upgrades made to each device, and results from repeatability and comparison testing between the two devices.

Abstract:
State highway agencies (SHAs) recognize that the public judges pavements and bridges by their smoothness and noise. This has resulted in most SHAs using smoothness pay incentives/disincentives for both concrete and asphalt pavements and some SHAs also include them for bridges. As SHA staffing has decreased over time, collecting the profile, computing the smoothness index of choice, and calculating the pay factor for the pavement or bridge have shifted from the SHA to the contractor. This started during the era of hand-pushed profilographs and has continued through to the high-speed profilers used today.
Federal regulations require that if contractor test data is used in the acceptance decision, the quality of the material has been validated by the verification sampling and testing on samples that are taken independently. Regardless of the funding source, SHAs are uncertain of the amount of independent data to collect and analyze to ensure they are making the correct acceptance decision and expending public funding correctly. This decision needs to be both risk based and cost effective as well as recognizing staffing restraints of SHAs. Other considerations could include:
• Evaluation of buyer’s and seller’s risk.
• Timing of the SHA collected profile versus that of the contractor, has the pavement or bridge profile changed due to the action of climate and traffic?
• Verification of contractors or vendors on a systematic rather than a project basis?
The rapid advancement of artificial intelligence (AI) tools presents both opportunities and risks for the pavement and bridge profiling verification processes. While AI can enhance efficiency and accuracy in data collection and analysis, it also introduces significant vulnerabilities, including the potential for data manipulation.
This presentation will discuss some common SHA practices for profile verification and the need for research to provide SHAs with cost-effective practices to verify contractor-collected profiles