Weigh-in-Motion (WIM) sensors are installed on mainline lanes at highway locations to record vehicle weights, axle spacing, vehicle class, travel speed, vehicle length, and traffic volume. These data elements support effective transportation planning, infrastructure design, and policy development. Therefore, it is important that WIM sensors supply accurate data. After initial installation and calibration, WIM systems may experience measurement drifts in weight and axle detection. Recalibration takes two general forms: (a) On-site calibration involving running trucks of known weight over WIM scales and (b) Auto-calibration methods involving comparisons to assumed reference weights. Auto-calibration can be more cost and time effective than on-site calibration. This paper leverages the increasing prevalence of truck tracking technologies like Global Positioning Systems (GPS) to improve auto-calibration methods and was divided into three aims: (i) data collection, (ii) data processing and (iii) model development. Truck GPS data from a national provider, WIM recorded truck weights, and static weights collected at weight enforcement station were gathered at several highway locations in Arkansas. A “matching” algorithm was developed to automatically match each GPS record to a WIM record based on timestamp and vehicle configuration. Algorithm performance was assessed via manual video verification of matches. Approximately, 75% of WIM and truck GPS records were correctly paired. Lastly, an auto-calibration model was developed to estimate lane and site specific calibration factors. The algorithm estimates hourly calibration factors by comparing the front axle weight of the same truck as it passes multiple WIM sites. Algorithm performance was measured by comparing estimated front axle and gross vehicle weights to known weights of the same truck measured at a static enforcement scale. The algorithm achieved Median Absolute Percent Error (MdAPE) of 11-23% for front axle weight and 15-45% for gross vehicle weight. These results can be improved by increasing the number of trucks that are able to be tracked across WIM sites with Automatic Vehicle Identification.
","abstract_html":"<p>Weigh-in-Motion (WIM) sensors are installed on mainline lanes at highway locations to record vehicle weights, axle spacing, vehicle class, travel speed, vehicle length, and traffic volume. These data elements support effective transportation planning, infrastructure design, and policy development. Therefore, it is important that WIM sensors supply accurate data. After initial installation and calibration, WIM systems may experience measurement drifts in weight and axle detection. Recalibration takes two general forms: (a) On-site calibration involving running trucks of known weight over WIM scales and (b) Auto-calibration methods involving comparisons to assumed reference weights. Auto-calibration can be more cost and time effective than on-site calibration. This paper leverages the increasing prevalence of truck tracking technologies like Global Positioning Systems (GPS) to improve auto-calibration methods and was divided into three aims: (i) data collection, (ii) data processing and (iii) model development. Truck GPS data from a national provider, WIM recorded truck weights, and static weights collected at weight enforcement station were gathered at several highway locations in Arkansas. A “matching” algorithm was developed to automatically match each GPS record to a WIM record based on timestamp and vehicle configuration. Algorithm performance was assessed via manual video verification of matches. Approximately, 75% of WIM and truck GPS records were correctly paired. Lastly, an auto-calibration model was developed to estimate lane and site specific calibration factors. The algorithm estimates hourly calibration factors by comparing the front axle weight of the same truck as it passes multiple WIM sites. Algorithm performance was measured by comparing estimated front axle and gross vehicle weights to known weights of the same truck measured at a static enforcement scale. The algorithm achieved Median Absolute Percent Error (MdAPE) of 11-23% for front axle weight and 15-45% for gross vehicle weight. These results can be improved by increasing the number of trucks that are able to be tracked across WIM sites with Automatic Vehicle Identification. </p>","abstract_has_math":false,"creators":["Zhang Durandal, Fu Ren Zhang"],"institution":null,"degree_name":"Master of Science in Civil Engineering (MSCE)","degree_level":"Thesis","degree_discipline":null,"degree_department":null,"school":null,"contributors":["Wood, Clinton M.","Prinz, Gary S."],"advisors":["Hernandez, Sarah V."],"committee_chairs":[],"committee_members":[],"year":2019,"date_issued":"2019-12-01T08:00:00Z","date_published":"2019-12-01T08:00:00Z","updated_at":"2026-07-24T00:59:40Z","subjects":["Auto-Calibration","Automatic Vehicle Identification","Weigh in Motion","Civil Engineering","Transportation Engineering"],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://scholarworks.uark.edu/etd/3463","outbound_label":"Repository record","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Wood, Clinton M.","Prinz, Gary S."]},{"key":"dc:contributor.advisor","label":"Advisor","values":["Hernandez, Sarah V."]},{"key":"dc:creator","label":"Author","values":["Zhang Durandal, Fu Ren Zhang"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2019"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2024-02-06T08:00:00Z"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Master of Science in Civil Engineering (MSCE)"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Auto-Calibration","Automatic Vehicle Identification","Weigh in Motion","Civil Engineering","Transportation Engineering"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://scholarworks.uark.edu/etd/3463"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Weigh-in-Motion (WIM) sensors are installed on mainline lanes at highway locations to record vehicle weights, axle spacing, vehicle class, travel speed, vehicle length, and traffic volume. These data elements support effective transportation planning, infrastructure design, and policy development. Therefore, it is important that WIM sensors supply accurate data. After initial installation and calibration, WIM systems may experience measurement drifts in weight and axle detection. Recalibration takes two general forms: (a) On-site calibration involving running trucks of known weight over WIM scales and (b) Auto-calibration methods involving comparisons to assumed reference weights. Auto-calibration can be more cost and time effective than on-site calibration. This paper leverages the increasing prevalence of truck tracking technologies like Global Positioning Systems (GPS) to improve auto-calibration methods and was divided into three aims: (i) data collection, (ii) data processing and (iii) model development. Truck GPS data from a national provider, WIM recorded truck weights, and static weights collected at weight enforcement station were gathered at several highway locations in Arkansas. A “matching” algorithm was developed to automatically match each GPS record to a WIM record based on timestamp and vehicle configuration. Algorithm performance was assessed via manual video verification of matches. Approximately, 75% of WIM and truck GPS records were correctly paired. Lastly, an auto-calibration model was developed to estimate lane and site specific calibration factors. The algorithm estimates hourly calibration factors by comparing the front axle weight of the same truck as it passes multiple WIM sites. Algorithm performance was measured by comparing estimated front axle and gross vehicle weights to known weights of the same truck measured at a static enforcement scale. The algorithm achieved Median Absolute Percent Error (MdAPE) of 11-23% for front axle weight and 15-45% for gross vehicle weight. These results can be improved by increasing the number of trucks that are able to be tracked across WIM sites with Automatic Vehicle Identification.
"]},{"key":"dc:title","label":"Title","values":["Weigh-in-Motion Auto-Calibration Using Automatic Vehicle Identification"]}]}],"canonical_facts":{"dc:contributor":["Wood, Clinton M.","Prinz, Gary S."],"dc:contributor.advisor":["Hernandez, Sarah V."],"dc:creator":["Zhang Durandal, Fu Ren Zhang"],"dc:date":["2019"],"dc:date.available":["2024-02-06T08:00:00Z"],"dc:description.abstract":["Weigh-in-Motion (WIM) sensors are installed on mainline lanes at highway locations to record vehicle weights, axle spacing, vehicle class, travel speed, vehicle length, and traffic volume. These data elements support effective transportation planning, infrastructure design, and policy development. Therefore, it is important that WIM sensors supply accurate data. After initial installation and calibration, WIM systems may experience measurement drifts in weight and axle detection. Recalibration takes two general forms: (a) On-site calibration involving running trucks of known weight over WIM scales and (b) Auto-calibration methods involving comparisons to assumed reference weights. Auto-calibration can be more cost and time effective than on-site calibration. This paper leverages the increasing prevalence of truck tracking technologies like Global Positioning Systems (GPS) to improve auto-calibration methods and was divided into three aims: (i) data collection, (ii) data processing and (iii) model development. Truck GPS data from a national provider, WIM recorded truck weights, and static weights collected at weight enforcement station were gathered at several highway locations in Arkansas. A “matching” algorithm was developed to automatically match each GPS record to a WIM record based on timestamp and vehicle configuration. Algorithm performance was assessed via manual video verification of matches. Approximately, 75% of WIM and truck GPS records were correctly paired. Lastly, an auto-calibration model was developed to estimate lane and site specific calibration factors. The algorithm estimates hourly calibration factors by comparing the front axle weight of the same truck as it passes multiple WIM sites. Algorithm performance was measured by comparing estimated front axle and gross vehicle weights to known weights of the same truck measured at a static enforcement scale. The algorithm achieved Median Absolute Percent Error (MdAPE) of 11-23% for front axle weight and 15-45% for gross vehicle weight. These results can be improved by increasing the number of trucks that are able to be tracked across WIM sites with Automatic Vehicle Identification.
"],"dc:identifier":["https://scholarworks.uark.edu/etd/3463"],"dc:subject":["Auto-Calibration","Automatic Vehicle Identification","Weigh in Motion","Civil Engineering","Transportation Engineering"],"dc:title":["Weigh-in-Motion Auto-Calibration Using Automatic Vehicle Identification"],"thesis:degree_level":["Thesis"],"thesis:degree_name":["Master of Science in Civil Engineering (MSCE)"]},"updated_at":"2026-07-24T00:59:40Z"}