Figure 9a shows the projection of rebars to the XZ plane. The estimated circle based on the Pratt algorithm is shown in Figure 9b. Table 2 shows the diameter of different sizes of the rebar in inch and mm units. The parameters such as light, camera angle of data collection, distance, selected resolution, and FOV will affect the final result of RGBD camera 3D point cloud data. Those parameters show their effect on how the actual object and the generated 3D point cloud data differ. So, there is an error that can be representative of these effects. To consider all the sources of errors in the proposed algorithm and simplify it for the first step of using an efficient low-cost camera in automated inspection, 50% less or more than any rebar size is considered as the acceptable range for algorithm to estimate the diameter of that size. Table 3 demonstrates the output of the algorithm for each rebar in three experiments. The results show that for the three experiments, all rebar diameters are correctly estimated by using the 50% methodology explained earlier. Experiment 1 and Experiment 2 have one #3 bar located automatically in positions 2 and 4, respectively. Experiment 3 has all #3 bars located correctly. Although this algorithm was able to identify the rebar #3 location correctly, there is an error in estimating the exact diameter of the rebar in different positions of each experiment. Equation 1 represents how the error is calculated. ​ ​ (1)​ Error(%)​ ​ = ​ ​ (_​)​​​reald​−​​​​estd​ d​real​​ ​ * 100​ where dest is the algorithm diameter estimation, and dreal is the diameter of the rebar based on its real size. Table 4 shows the error of the diameter estimation algo- rithm for each experiment. The average error of the algorithm is 6.47%. So, if the inspector wants to use the RGBD camera to detect the specific size of rebar and estimate the diameter, that can be a helpful way with an estimation of almost 7% error. This experiment was an example of how the RGBD camera can be used to create the 3D point cloud data of the structure and acquire geometric information of the rebar, which is nec- essary for inspecting the structure. The density of the rebar, the background, and the environmental condition will impact ME | RGBDCAMERAS Figure 9. Results of experiment: (a) rebar projection on XZ plane (b) estimated fitted circle. Experiment 1 Experiment 2 Experiment 3 #3 #3 #3 #3 #3 #3 TA B L E 2 Rebar diameter and acceptable range for algorithm Bar size Nominal diameter (in.) Nominal diameter (mm) Algorithm acceptable range for each size (mm) Minimum Max #3 0.375 9.525 7.938 11.113 #4 0.500 12.700 11.113 14.288 #5 0.625 15.875 14.288 17.463 #6 0.750 19.050 17.463 20.638 #7 0.875 22.225 20.638 23.813 #8 1.000 25.400 23.813 27.025 TA B L E 3 Diameter estimation for three experiments Estimated diameter (mm) Position 1 Position 2 Position 3 Position 4 Experiment 1 11.9 8.8 12.4 11.6 Estimated size #4 #3 #4 #4 Experiment 2 13.5 13.2 14.2 9.3 Estimated size #4 #4 #4 #3 Experiment 3 9.9 9.8 10.2 10.6 Estimated size #3 #3 #3 #3 TA B L E 4 Diameter estimation error for three experiments Error (%) Position 1 Position 2 Position 3 Position 4 Experiment 1 -5.96 -7.43 -1.96 -8.55 Experiment 2 6.50 4.55 12.35 -2.17 Experiment 3 4.83 3.56 7.79 12.03 52 M AT E R I A L S E V A L U AT I O N • J A N U A R Y 2 0 2 3 2301 ME Jan New.indd 52 12/20/22 8:15 AM

the result of the recorded data from the camera. It makes the process more complicated in terms of computational time and effort. However, the preprocessing step in CloudCompare will help to remove the background and divide the top and bottom layer based on the Z elevation. Eventually, implementing the algorithm on the refined data can be done. Further Research The results show that the RGBD camera can be a low-cost alternative to enhance construction inspectors’ capabilities in the global assessment of RC concrete construction, spe- cifically rebar size distinction. The future focus of research is to incorporate RGBD cameras with AI techniques to develop tools to assist the construction inspectors in (a) checking size, number, location, and spacing against shop drawings and plans (b) checking rebar for proper clearance between bars, joints, and side forms (c) checking the adequacy of ties and tie downs (d) checking splice locations and lengths against shop drawings and plans (e) assuring bars with suitable coating are being used as noted in the plan (f) detecting coating deficien- cies and (g) recording the placement of rebar in daily reports submitted by the inspector. If successful, the same principles can be extended to assist the construction inspector with post-placement of rebar activities, including assuring the rebar is not displaced during concrete placement and the proper cover is maintained. The final product will be to generate a report of deficiencies on-site so that they can be addressed in a timely fashion to promote durability while avoiding construc- tion delays. Additionally, the digital as-built model showing the rebar location and size can be used to report and record the progress of rebar construction in real time. Requirements on rebar spacing can also be incorporated into the process as an additional measure to assure design drawings meet the appli- cable specifications (such as AASHTO and ACI). This research also has potential for other applications, such as changes in safety equipment and obstructions/depressions in the con- struction area to increase worker safety, as well as verification of pipeline/bridge layouts. Another application that can be considered for the RGBD camera is to combine it with other technologies such as UAVs. The RGBD camera can be mounted on the UAV, which can help the inspector to collect data from a large area in a short time. It should be noted that the camera needs to be mounted on the drone, so a 3D printed setup or gimbal can be used to reduce the vibration and subsequent noises in the data. Then, the inspector would be able to have a digital as-built model of the construction site. Additionally, the inspector can use AR technology and wear a headset on-site to see the real-time 3D data created by the RGBD camera in front of their eyes as it shows which rebar should be replaced in holograms. Also, in the future, the authors would like to compare the results from RGBD and LiDAR, as each has its own advantages. The RGBD camera’s application is not limited to the structural construc- tion site it can also be utilized in quality control of under- ground utilities as the inspector crew can create a 3D point cloud data model of the underground facilities to detect the defects. Conclusion The RGBD camera used in this study has the capability of generating 3D point cloud data of the structure with different features and modes based on the user-specific application. The authors presented a method of using this camera to rec- ognize rebar #3 between rebar of different sizes using the Pratt circle fitting algorithm to estimate the diameter. The results demonstrated that the developed algorithm can automati- cally and quickly acquire measurements comparable to a tape measure utilizing the inexpensive handheld RGBD camera. This is useful for the inspector to distinguish between rebar of similar sizes, as it is hard to differentiate between them from a distance with human eyes. This paper also explained the importance of using tech- nologies for automation in the quantification of inspection. It decreases the amount of time required for manually recording data and has the capability of creating permanent construc- tion records. It also lets the inspectors monitor the whole structure’s state precisely and comprehensively and creates a large amount of data that could be a source of information and comparison for the next inspection. The results showed that the RGBD camera holds promise as a low-cost alternative to enhance construction inspectors’ capabilities in the global assessment of RC concrete construction, specifically rebar size distinction. ACKNOWLEDGMENTS The authors of this paper acknowledge the support from the Department of Civil, Construction, and Environmental Engineering of the University of New Mexico the Transportation Consortium of South-Central States (Tran-SET, Award Number 69A355 l 747106 Grant ID: 2RTR1-2TR13) the Federal Railroad Administration (FRA) Award No. 693JJ621C000010 and the New Mexico Department of Transportation (NMDOT) for providing funding to graduate students to conduct this research. We appreciate the input and assistance from Ronan Reza. REFERENCES Akinci, B., F. Boukamp, C. Gordon, D. Huber, C. Lyons, and K. Park. 2006. “A formalism for utilization of sensor systems and integrated project models for active construction quality control.” Automation in Construc- tion 15 (2): 124–38. https://doi.org/10.1016/j.autcon.2005.01.008. Almadhoun, R., T. Taha, L. Seneviratne, J. Dias, and G. Cai. 2016. “A survey on inspecting structures using robotic systems.” International Journal of Advanced Robotic Systems 13 (6). https://doi.org/10.1177/1729881416663664. Ayele, Y. Z., M. Aliyari, D. Griffiths, and E. L. Droguett. 2020. “Automatic crack segmentation for UAV-assisted bridge inspection.” Energies 13 (23): 6250. https://doi.org/10.3390/en13236250. Biswas, K.K., and S.K. Basu. 2011. “Gesture recognition using Microsoft Kinect®.” The 5th International Conference on Automation, Robotics and Applications, Wellington, New Zealand. https://doi.org/10.1109/ ICARA.2011.6144864. Boukamp, F., and B. Akinci. 2007. “Automated processing of construction specifications to support inspection and quality control.” Automation in Construction 17 (1): 90–106. https://doi.org/10.1016/j.autcon.2007.03.002. BrÜhwiler, E. 2009. “Improving Safety and Durability of Civil Structures.” In Damage Assessment and Reconstruction after War or Natural Disaster, eds A. Ibrahimbegovic and M. Zlatar. NATO Science for Peace and Security Series C: Environmental Security. Dordrecht: Springer. https://doi. org/10.1007/978-90-481-2386-5_3. J A N U A R Y 2 0 2 3 • M AT E R I A L S E V A L U AT I O N 53 2301 ME Jan New.indd 53 12/20/22 8:15 AM