Network Level Crack Survey with the Automated Real-Time Distress Analyzer KELVIN C.P. WANG 1, CURTIS NUNN 2, CHAD MACKEY 2, WEIGUO GONG 2, DAVID WILSON 2, MARK EVANS 3, AND JERRY DALEIDEN 4 TOTAL NUMBER OF WORDS, 7,000, INCLUDING FIGURES AND TABLES ABSTRACT The Arkansas State Highway and Transportation Department (AHTD) contracted the University of Arkansas to conduct a network crack survey of a large portion of its noninterstate National Highway System (NHS). The Digital Highway Data Vehicle (DHDV) was used to acquire high-resolution digital images and analyze cracks with the automated real-time Distress Analyzer. This paper presents a preliminary study using the DHDV and the Distress Analyzer for data analysis on a network of about 100 miles of pavements. In addition, a manual survey was conducted on the same network of pavements. The data analysis with the Distress Analyzer covers the entire network, while the manual survey covers 5% of the same area on a mile-by-mile basis. Three distress protocols were used in the data analysis: the AASHTO Interim Distress Protocol, the Texas DOT Method, and World Bank s Universal Cracking Indicator (CI). The study demonstrates that the 1 4190 Bell Engineering, Civil Engineering, University of Arkansas, Fayetteville, AR 72701, USA, kcw@engr.uark.edu, Phone: 501-575-8425, Fax: 501-575-7168 2 University of Arkansas, Fayetteville, AR 72701 3 Research and Planning, Arkansas State Highway and Transportation Department, Little Rock 4 Fugro-BRE, Austin, Texas 1
automated real-time Distress Analyzer is effective in speed and accuracy. Because the Distress Analyzer is fully automated and results of the analysis are provided simultaneously to image collection, the potential cost savings when compared with manual survey methods and other semi-automated survey technologies is tremendous. The study also reveals that image quality is a critical factor for the Distress Analyzer and additional features need to be incorporated into the analyzer, such as more accurately determining lane markings on both sides. Due to limitation of time and space, only results based on Universal Cracking Indicator (CI) is presented in the paper. INTRODUCTION The Digital Highway Data Vehicle (DHDV) is the first fully automated real-time distress survey system designed for crack detection. The basic capabilities of the Distress Analyzer as incorporated into the DHDV include real-time analysis of pavement images for cracks, creation of crack maps with geometric crack information, and storage of cracking information in database tables. Report Writer software can then be applied to the database tables to summarize the data into distress indices. Wang et al (1) published a summary of distress survey system designs and performance over the last two decades. In 2002, Wang (2) presented an initial study on the use of the Distress Analyzer and Report Writer for sample sections. The same publication also presented a detailed description of the general capabilities of the DHDV and the Distress Analyzer, which are not addressed in this paper. Figure 1 shows a screen shot of the distress analyzer working at real-time with two Pentium-III processors at 733-MHz per CPU. As can be seen from Figure 1, two analyzers were launched in parallel with combined processing speed over 40 MPH with 2
corresponding two side-by-side processes. At the bottom of both processes are the analysis results for crack geometry and basic classification. Each process shows the original image on the left and processed binary image on the right. Each binary image shows identified cracks in bounding boxes with a unique integer number for each bounding box. A rectangular-shaped bounding box contains one crack. The right-most window illustrates the processing status of the two analyzers. In 2002, the Arkansas State Highway and Transportation Department (AHTD) contracted the University of Arkansas to conduct a distress survey of a large portion of its non-interstate Nation Highway System (NHS). This paper presents preliminary data findings on about 100 miles of US 412 pavement across four adjacent counties in northern Arkansas (Figures 2 and 3). Three distress protocols are incorporated into the Report Writer, a software program that examines geometric crack information and summarizes data into protocol indices. The three protocols include the AASHTO Interim Distress Protocol, the Texas DOT Distress Method, and World Bank s Universal Cracking Indicator (UI). SURVEY PROTOCOLS Protocol Descriptions AASHTO PROTOCOL The AASHTO interim distress protocol (3) quantifies asphalt pavement surface cracking based on crack severity within wheelpaths, between wheelpaths, and outside wheelpaths. This standard is designed primarily for automated equipment and is best suited for 100% roadway coverage. The AASHTO protocol defines a crack as a 3
discontinuity in the pavement surface with minimum dimensions of 3 mm (1/8 in) width and 25 mm (1 in) length. Cracks may include longitudinal cracks, transverse cracks, and interconnected (fatigue and block) cracks. The AASHTO protocol quantifies the cracks and differentiates between load associated (fatigue) and non-load associated (environmental, reflective, etc.) pavement cracking and joints. Increased cracking intensity in the wheelpath as compared to the non-wheelpath areas is assumed to quantify load associated cracking. Non-load associated cracking is quantified by the cracking measured in the non-wheelpath areas. Wheelpath cracking is determined in both the inside and outside wheelpath. Non-wheel path cracking is determined for the area between the wheelpaths and the area outside the wheelpaths. Illustration of wheelpath and nonwheelpath is shown in Figure 4. The AASHTO method defines crack severity levels as follows: Severity Level 1: Cracks smaller than 3 mm (1/8 in) Severity Level 2: Cracks with widths from 3 mm (1/8 in) to 6 mm (1/4 in) Severity Level 3: Cracks with widths greater than 6 mm (1/4 in) Each cracking level is quantified by the total length of cracking per unit area (m/m 2 ) for each defined survey strip. In the experiment we conducted, wheel and non-wheel path areas are determined by using the measurements illustrated in Figure 4 from the Long Term Pavement Performance (LTPP) Project Distress Manual (4). 4
UNIVERSAL CRACKING INDICATOR The Universal Cracking indicator (CI) (5) is the simple product of the three primary physical dimensions observed for cracks: CI = extent intensity crack width... (1) where extent is the area of cracked pavement defined within a sample area, which may be the perimeter bounding a set of cracks, expressed as a percentage of total pavement area. The intensity is the total length of cracks within the area defining the extent (expressed in m/m 2 ). The crack width is the mean width of crack opening at the surface of a set of cracks (expressed in mm). Figure 5 illustrates the basic graphical concept for Universal Crack Indicator (CI). For longitudinal cracking, the CI is calculated as a ll b 0 100lLw CI L = 100 wb + 0 = A a A b A L... (2) For alligator cracking, the CI is calculated as C l A 100l Aw CI A = 100 wa = A C A A... (3) For transverse cracking, the CI is calculated as A lt lt wt CI T = 100 wt = 100 A A A... (4) The final CI is the sum of the above three indicators: [ w + l w l w ] A CI L + = 100 l L A A T T /... (5) 5
TEXAS DOT PROTOCOL Unlike the two methods above, which calculate indicators disregarding crack types, the Texas DOT method (6) gives a quantifying value for several different types of cracks. The method is described in detail in the Texas DOT Pavement Management Information System (PMIS) Rater s Manual (6). This survey dealt with four types of cracks. Longitudinal cracking was measured in terms of linear feet per station (i.e. feet of cracking in each 100 feet (31 meters) of surface). Transverse cracking was measured in terms of the number of full lane-width cracks per station (i.e. number of cracks in each 100 feet (31 meters) of surface). The rating value for alligator (fatigue) cracking measures the percentage of the rated lane s total wheel-path area that is covered by alligator cracking. Block cracking values were measured as a percentage of sample lane area covered by full lane-width block cracking. Texas DOT rating methods for alligator and block cracks are similar to the extent calculations of the CI method. Protocol Implementations For the 100-mile network level survey, both the manual rating and Distress Analyzer systems used exactly the same set of digital images to conduct pavement distress analyses. The Distress Analyzer rated 100% of the project road surface images and determined cracking criteria values by actual measurement of discontinuities. Data was stored in Microsoft Access database files that were then examined by Report Writer software that calculated protocol results by section and output them in a Microsoft Excel spreadsheet format. Manual survey results were based upon visual examination of images 6
of the first 260 feet of each mile (5% of pavements) viewed on a computer monitor at resolution of 1600 x 1200. The decision to conduct survey on only 5% of pavements is due to limitation of availability of raters and time. Estimated rating values were recorded manually onto hardcopy worksheets (Figure 5) and later transferred to Microsoft Excel spreadsheets where custom functions were developed for data and statistical analysis. Results from manual surveys are assumed to be representative of the entire mile. The first mile of each county was omitted for both eh Distress Analyzer and manual survey from the study due to tuning image quality during the initial stages of image capture that were not conducive to distress analysis. AASHTO PROTOCOL Crack length per square meter for severity levels 1-3 were reported for each wheelpath position. For longitudinal and transverse cracks, the total estimated length of cracking per 20-foot division as determined by the manual survey was subdivided equally among all severity levels and all wheelpath positions noted for that division. The Distress Analyzer performed more precise divisions of cracking among multiple wheelpath positions and severity levels based on actual measurement. Both methods report fatigue cracking only in wheelpaths; therefore, fatigue cracking was only divided among severity levels. Block cracking was subdivided among all severity levels and wheelpath positions outside and between wheelpaths. UNIVERSAL CRACKING INDICATOR To determine extent values, the manual survey added the estimated areas of all cracks in the 260-foot sample area for each mile. Widths of 1 foot were arbitrarily assigned to 7
longitudinal and transverse cracks, due to the fact that manually determining crack width for extent values would take too much time for the rater and would lack necessary precision anyhow. In addition, as the unit of measurement is in foot, the result for extent values is actually the summation of lengths of cracks over total area. All fatigue cracks were given widths of 2.49 feet (0.76 meters), the width of a wheelpath. Block cracks were assumed to be equidimensional, and their longitudinal length was squared to calculate area. The summed total of all crack areas were then divided by the total area of the 260-foot sample zone. The Distress Analyzer performed more accurate quantification of crack areas by using precise measurement capabilities within the software. Both manual survey and Distress Analyzer intensity values were calculated by summing cracking data for all wheelpath positions in all severity levels as outlined in the AASHTO protocol. Crack width was calculated for the manual survey by multiplying the crack length of each type of crack by its average crack width in each 20-foot division then dividing that product by the total crack length for each division. Table 1 correlates marked severity levels and their corresponding average crack width. The Distress Analyzer used actual measurements to average crack widths for each section. TEXAS DOT PROTOCOL To manually determine length in feet of longitudinal cracking per station for each mile, a rater estimated the total length of all crack types in each 20-foot division of the first 200 feet of the mile. These lengths were summed and halved to calculate an average length of longitudinal cracking per100-foot station for each section. The Distress Analyzer 8
calculated measured lengths of longitudinal cracking per station for each section. Section values were averaged to obtain cracking per 100-foot station for each section. The number of full lane-width transverse cracks per station for each section was determined manually by estimating the total length of transverse cracks in each 20-foot division of the first 200 feet of the mile. These lengths were summed and halved to calculate an average length of transverse cracking per 100-foot station for each section. That value was then divided by 12 to obtain an average number of full lane-width transverse cracks per station per section. The Distress Analyzer averaged section values for calculated number of transverse cracking per station to obtain number of full lane-width transverse cracks per station for each section. The percent of wheelpath area covered by fatigue cracking was calculated manually by summing all 20-foot division estimated fatigue crack lengths to obtain the total length of fatigue cracking over the first 260 feet of each mile and then dividing that sum by 520, the combined length of two wheelpaths through the 260 foot length. The Distress Analyzer determined the actual percentage of section covered by fatigue by obtaining measurements of crack area per section and dividing by the section area. The percent of road surface area covered by full lane-width block cracking was calculated manually by squaring the total estimated block crack length for each 20-foot division of the first 260 feet of each mile. The sum of those products was then divided by the area of the 260 feet of road surface. No more than one block crack was observed for a single 20-foot division, so errors arising from squared combined block crack lengths did not occur. The Distress Analyzer determined the actual percentage of section covered by 9
fatigue by obtaining measurements of crack area per section and dividing by the section area. DATA ANALYSIS AND RESULTS Data Collection TEST SECTION Prior to the data collection of the network of pavements, seven consecutive sections of severely distressed pavement within the Fayetteville, Arkansas, city limits were chosen as a test section. Two manual raters and the Distress Analyzer inspected the entire seven-tenths of a mile in order to make direct comparisons over the section. This exercise was necessary to get the raters to be familiar with the distress protocols and the application of the Distress Analyzer. NHS NETWORK To compare manual and Distress Analyzer systems on a network level, a total of 100 miles of US 412 across four adjacent northern Arkansas counties (Figures 1 and 2) were surveyed by both rating techniques. Data Analysis Microsoft Excel spreadsheet data compilations output by the Report Writer and manually entered by the rater were compared in Excel via standard deviation analyses. Due to limitation of time to compose data for other indices, particularly the AASHTO protocol, CI values for the first section of each mile and the first mile as a whole were used in the analyses. Figures 7 to 10 and Tables 2 and 3 illustrate the results from the use of CI as the index for cracking survey of the network. Particularly, Tables 2 and 3 present data 10
for the 100 miles of pavements in the four counties: Madison, Carroll, Boone, and Marion. Table 2 includes results based on 5% manual survey and 100% automated survey (the entire pavements). Table 3 shows results based on 5% manual survey and 10% automated survey. The 10% pavements used in the automated survey includes 5% pavement sections used in manual survey. The reason for using 10% automated survey is because the current Report Writer does not have the capability to produce data at that interval. Due to limitation of time and space, and the convenience of using CI as a single index, we determined that discussion of CI only in this paper is representative. After compilation of all automated and manual survey data, results from the two methods were compared using a standard deviation statistical analysis. Where standard deviation between manual survey and Distress Analyzer data was high, both rating methods were reviewed in an attempt to isolate and correct the cause of the discrepancy. Based on the standard deviation values from the two tables, both analysis approaches generate similar results, even though the overall value for standard deviation shown in Table 2 is smaller than that in Table 3. This is expected and reasonable, since the 5% survey is more representative of the 10% pavements than the entire mile. Figure 7 to 10 graphically present data form surveys conducted in Madison and Carroll Counties from Tables 2 and 3. Each pair of images uses data from the two approaches. It can be seen from the graphs that when cracking presence on pavement sections is limited or does not exist, results from both manual survey and the Distress Analyzer are very close. For a large number of sections where cracking presence is substantial, differences between results from manual survey and results from the Distress 11
Analyzer do exist. But, in most cases the results from the two surveys are statistically similar, particularly when cracking presence is small to moderate. MADISON COUNTY In Figure 8, pavements on miles 9, 16, and 21 have an automated grade of well above the manual grade. This is due to the capture of cracks to the left of the near lane marker of the center double line. Figure 11 illustrates this problem. An adjustment to the lane size algorithm in the Distress Analyzer will create a better measure of the amount of cracking within the lane. This work is underway and will be completed shortly. In Figure 7, comparison of data from the Distress Analyzer over each entire mile versus results from the 5% manual survey, the results are again accurate. In Figure 7, for pavements on miles 13, 15, 23, and 24, cracking indices from the Distress Analyzer are much higher than the results from the manual survey. In miles 13, 15, and 24, it was found out by examining all relevant images that there exists large number of dense areas of cracking after the first 10% of each mile. This is particularly true for miles 13 and 15. In mile 23 of Figure 7, there exist a large number of cracks over the 2 nd 5% of the mile, therefore the manual survey could not catch the cracks, while the Distress Analyzer did. Overall, the Distress Analyzer produces higher cracking index values than the manual survey for most of the pavement sections on Madison 412. There are two reasons for this. The first is that the automated procedure may have found more cracks because it grades 100% of each mile. The second reason is the Distress Analyzer picked up some noises for cracks, such as stains, tire tracks, train rails. The Distress Analyzer ignores most noises. 12
When the image quality is not of the highest, it may occasionally making errors of including the noises. CARROLL COUNTY In Figure 10, pavements on miles 5, 9, 10, 14, and 17, have higher cracking indices from the Distress Analyzer than those from the manual survey. One reason for this discrepancy is the capturing and recording of cracking outside the lane markers, similar to the problem in Figure 11. Particularly for mile 9, the vehicle strayed off during this section of pavement from the center of the lane, causing the Distress Analyzer to recognize substantial amount of cracks outside of lane markings. Miles 5, 10, 14, and 17 all have substantial cracking beyond the first 5% of pavement, but within the 2 nd 5%, thus the higher cracking index. In Figure 9, data from the Distress Analyzer on entire miles are more representative of the results from manual survey than data from Figure 10. However, some discrepancies are present. Miles 4 and 7 have some image quality problems. Miles 18 and 19 have moderate cracking indices for the entire mile but do not have recorded values from the rater. Again, this is due to cracking occurring in the rest of mile, but not during the 1 st 10%. All other miles have accurate results from both survey methods. Experiences with the Distress Analyzer The Distress Analyzer does offer several advantages over manual distress surveys: 1. Distress Analyzer results are based on inspection of 100% of the study area. Manual surveys actually inspect only 5% of every mile and assume that that 5% is representative of the entire mile. Underestimation or overestimation of distress based 13
on interpolation of non- representative manual survey samples are avoided using the Distress Analyzer. 2. Distress Analyzer results are based on actual measurements. All distress data reported by the Distress Analyzer is based on empirical measurements. Values for crack width, crack length, and crack area are all determined by actual measurement. Equivalent manual survey values are estimates based strictly on the rater s interpretation of pavement images. 3. Distress Analyzer results are accurate. Although results from the Distress Analyzer and manual surveys in this paper do not agree perfectly, tabulated data comparisons show strong correlation between the overall distress character of specific pavement areas. As seen in the manual survey of the demonstration section, CI values produced from inspection of two different raters may show significant standard deviation, and the standard deviation associated with Distress Analyzer is acceptable when in light of this comparison. 4. Distress Analyzer results are consistent. As demonstrated by Wang and others (1), Distress Analyzer results for multiple passes over the same sample are consistent to within about 15%. 5. The Distress Analyzer eliminates human error. There are several opportunities for human error in the manual rating process. These include errors estimating crack dimensions from viewed images, recording data on the survey chart, and transferring data from the chart to the Excel spreadsheet. A totally automated system eliminates human errors, and Wang and others (1) have shown the Distress Analyzer to be very consistent in crack recognition with repeated surveys over the same study area. 14
6. Distress Analyzer results are fast and inexpensive. Distress Analyzer software can perform real-time inspection of collected digital images at normal highway speeds. This means that at the completion of a data collection run, that data will have already been analyzed for distress. Distress Analyzer software can also be run in an unsupervised lab environment at equivalent highway speeds if simultaneous collection and inspection are not desirable. Manual rating of digitally-collected images is laborintensive and time-consuming. For this survey of 100 miles, manual rating time was 24 hours with an additional 4 hours for data entry. After the initial investment in Distress Analyzer software, no addition expenses associated directly with the Distress Analyzer will be incurred. Manual surveys require extended man-hours to perform. CONCLUSION A new image acquisition system with transverse resolution of 2,600 pixels will soon replace the current system in the effort to establish a system capable of capturing cracks at 1 mm size. Automated network level survey for pavement cracks has been a main focus for researchers and practitioners in recent decades. Field deployment of such technologies has begun just recently. Even though there are several necessary improvements to be made to the DHDV and the Distress Analyzer developed at the University of Arkansas, the operating performance and efficiency of the system are clear. With an acquisition system with twice the resolution and further improvement of imaging algorithms, such as more accurately determining lane markings on both sides, it is anticipated that in the next few years that widespread use of fully automated crack survey systems will be a reality. 15
REFERENCES 1 Wang, K.C.P. Design and Implementation of Automated Systems for Pavement Surface Distress Survey, ASCE Journal of Infrastructure Systems, Vol.6, No1, March, pp. 24-32, 2000 2 Wang, K.C.P., Gong, W.G., Li, X.Y., Elliott R. P., and Daleiden, J. Data Analysis of A Real-Time System for Automated Distress Survey, in print for publication, Transportation Research Board, Washington, D.C., 2002 3 AASSHTO, Standard Practice for Quantifying Cracks in Asphalt Pavement Surface, AASHTO Designation: PP44-00, 2001 4 Strategic Highway Research Program National Research Council, Distress Identification Manual for the Long-Term Pavement Performance Project, 1993 5 William D. Paterson, Proposal of Universal Cracking Indicator for Pavements, Transportation Research Record 1455, pp. 69-74 6 Texas Department of Transportation, Pavement Management Information System Rater s Manual (For Fiscal Year 1999) 16
Figure 1 Dual-Processing for Cracks in a Parallel Environment with the Crack Analyzer 17
Figure 2 NHS distress survey area across Madison, Carroll, Boone, and Marion counties, from west to east. Figure 3 US 412 (light green) across distress survey counties totaling about 100 miles. Data was collected from west to east in the direction of increasing log miles. 18
Figure 4 Lane and wheelpath dimensions as set forth by LTPP. Figure 5 The Basic Graphical Concept for the Universal Crack Indicator. 19
Figure 6 Sample manual survey data sheet. 20
Figure 7 Cracking (CI) Results for US 412, Madison County, 5% Manual Survey Versus 100% Automated Survey Figure 8 Cracking (CI) Results for US 412, Madison County, 5% Manual Survey Versus 10% Automated Survey 21
Figure 9 Cracking (CI) Results for US 412, Carroll County, 5% Manual Survey Versus 100% Automated Survey Figure 10 Cracking (CI) Results for US 412, Carroll County, 5% Manual Survey Versus 10% Automated Survey 22
Figure 11 Cracking within and outside of double line picked up by analyzer, Mile 21, Hwy 412, Madison County 23
Table 1 Manual survey average crack width values assigned to recorded severity levels Recorded Severity Level(s) Average Crack Width Value (mm) 1 2 1 & 2 3.25 1 & 3 4.5 1, 2, & 3 4.5 2 4.5 2 & 3 5.75 3 7 24
Table 2 Universal Cracking Indicator (CI) Values, Manual Rating with 0.05-Mile, Distress Analyzer with Complete Coverage 25
Table 3 Universal Cracking Indicator (CI) Values, Manual Rating with 0.05-Mile, Distress Analyzer with 0.1-Mile Coverage 26