Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
27 From the survey data, five agencies were selected for follow-up interviews. These interviews were used to supplement information from the survey and gather additional insight and specific experiences. Using the survey as a screening tool, the research team could formulate questions specific to an agency and target areas of interest. From the list of agencies that indicated a willingness to participate in a follow-up interview, the following characteristics of interest were developed to narrow the field to the five DOTs: â¢ Geographic diversity â¢ Population diversity â in terms of the total population in a state â¢ Whether they collect data â to include a ânoâ perspective â¢ Diversity in how they use data â¢ Data clean-up, use in multiple applications, other criteria including freight â¢ Diversity in who responded (not all engineers/planners) â¢ Law enforcement or weight enforcement responsibility â¢ Freight nexus â¢ Response to questions on challenges with data â¢ Border with Canada and Mexico (where freight flows may differ from regular interstate flows) â¢ East/West/Gulf Coast/Great Lakes (to capture freight flows from international trade and border crossings) California, Minnesota, Florida, Maryland, and Tennessee DOTs were selected for case example interviews. Table 1 shows the selection of the five DOTs and some of the consider- ations for selection. Based on the initial survey responses, a series of interview questions was developed for each agency. The survey and follow-up interview formed the basis of the following case examples. California The California DOT (Caltrans) was selected for a follow-up interview as California is a large, populous state with ports, agriculture, and a border with Mexico. Below is a summary of the Caltrans response. Sensors and Systems Caltrans uses mostly bending plate WIM sensors in their more than 140 WIM sites. The main reason for using bending plate sensors is that, due to traffic congestion, in most locations Caltrans has restricted lane closure times. In most areas, Caltrans has an 8-hour lane closure C H A P T E R 4 Case Examples
28 Use of Weigh-in-Motion Data for Pavement, Bridge, Weight Enforcement, and Freight Logistics Applications limit and in a few places, lane closures are restricted to 5 hours. Caltrans has fine-tuned the sensor replacement process and been able to replace a bending plate sensor in as little as 45 minutes, not including the lane closing and reopening time. Caltrans has tried load cell sensors. These give better data and law enforcement likes them, but they take up to 3 to 4 days to install, considered an excessive amount of time. All WIM stations are installed in concrete. There is one quartz piezo-electric station in the asphalt pavement at the Port of Long Beach, but the port helped pay for this installation. Other sensors require more time and clean-up than Caltrans can allow. Caltrans has been approached by universities to study bridge WIM using cameras to monitor bridge member deflection. Caltrans has not done any of this themselves as the accuracy of this new technology is unknown. Caltrans monitors data for quality control (QC) and to determine if systems are calibrated adequately. Caltrans periodically has data transfer problems from the phone lines used to trans- fer data and sometimes the QC protocols break down. Phone lines are used because the expense is low compared with cellular telephone service. QC problems have arisen because of staff turn- over and a lack of training opportunities. Caltrans monitors front axle weights to look for drift because that signals a need for calibration; the front axle is least affected by the truckâs load. Caltrans is mostly satisfied with their WIM network, but sees some holes in the Oakland and Stockton areas. The biggest impediment to growing the system is the approximate $1.4 million cost for a four-lane WIM installation. Data Usage Caltrans has used AASHTO ME for concrete pavement design, developing a design catalog for concrete pavement; it is not really needed for design anymore because designs are deter- mined with basic project information. However, it has also been used for concrete pavement failure investigations and other research-related activities. For asphalt pavements, Caltrans has developed their own design software (CalME), which is based on the same principles of mechanistic-empirical design like the AASHTO ME. CalMEâs traffic relies on loading spectra models developed with WIM data. WIM data were used to revise the load factors on truck loads on a few selected bridges serving the ports in California, such as the new Port of Long Beach bridges. For design of typical bridges, WIM data have not been used. DOT Identified WIM Use/Interest from Survey Other Identified Selection Factors California Pavement ME Design, bridge design Large state, populous state, ports, agriculture, border with Mexico Minnesota Pavement Design, interested in freight planning Large state, Great Lakes access, international border with Canada Florida Pavement ME Design, freight taxonomy East and Gulf coasts, ports, freight flow due to agriculture Maryland Beginning implementation of Pavement ME Design Small state, East Coast, major port, highly urbanized, developed WIM program Tennessee Freight planning Freight crossroads, no water ports, no use of WIM, but planning to redevelop a WIM program Table 1. DOTs for case example interviews.
Case Examples 29 Caltrans currently has no VWS stations. Law enforcement and Caltrans personnel are interested in studying overweight trucks moving through the state. In the past, privacy issues have been raised about using VWS. Caltrans has a research project with the University of California at Irvine to study what data can be determined just by using inductive loops. So far, researchers are making progress in iden- tifying truck types. Because loops are much less costly than full WIM sites, this holds promise to expand system freight knowledge at low cost and consequently more sites. Minnesota Minnesota DOT (MnDOT) was selected for a follow-up interview based on meeting several of the initial selection criteria. Minnesota is a larger state with Great Lakes access and the international border with Canada. The survey and follow-up interview produced the following snapshot of MnDOTâs use of WIM. Sensors and Systems In the 1980s MnDOT used bending plate sensors, but there was an incident where a plate broke off and hit a car. No one was injured, but the state Attorney Generalâs Office recommended discontinuing the use of bending plate sensors. The state tried load cell sensors, but they required lane closures for winter and spring prepa- ration and they were very expensive to install. The load cells are favored by the State Patrol because they last longer (up to 20 years) and are more accurate, but the DOT determined them to be too expensive and disruptive to traffic to justify using them. MnDOT tested strain-gage sensors, but they did not work well on the test site MnDOT has at MnROAD (MnDOTâs test track for pavement evaluation). Ceramic piezo-electric sensors were also tried, but suffered from inaccuracy due to Minnesotaâs temperature variations. From experience and testing, MnDOT only uses one brand of quartz piezo-electric sensors for all WIM installations. They are easier to install, are accurate, and last 4 to 7 years. MnDOT has 23 WIM sites, with 13 that also include cameras. MnDOT WIM systems are not without their problems. Cost is the main issue. MnDOT initiated a study to estimate the costs for a two-lane installation. An automatic traffic recorder (ATR) costs $40,000 and a WIM system costs $140,000. ATR data can provide volume, speed, and classification information. WIM data can provide weight information as well as volume, speed, and classification information. Recent costs for WIMs have increased because new installations include cameras. MnDOT, with 23 WIM systems, would like to obtain greater coverage by installing a few more stations in strategic locations. However, MnDOT indicated that several Big Data companies can supply information that contains elements of the data MnDOT captures with WIM stations, and this is worth exploring in the future. Another issue with systems is calibration. MnROAD is MnDOTâs test track, similar to the AASHO Road Test, where MnDOT uses trucks of known weight to allow testing of various pave- ment configurations and materials. The MnDOT WIM group used to work with the MnROAD staff to use an MnROAD truck in order to calibrate the WIM systems. Recently, the ability to use these trucks was curtailed. MnDOT is looking internally for a truck or perhaps a contract for
30 Use of Weigh-in-Motion Data for Pavement, Bridge, Weight Enforcement, and Freight Logistics Applications calibration. Out of necessity, MnDOT is switching from a set calibration schedule to monitoring accuracy and calibrating only when needed. Data Usage Law enforcement agencies have direct access to MnDOTâs data. Law enforcement can access systems and cameras at any time and use the WIM stations to screen trucks, pulling over for static weighing any trucks that appear overweight. In turn, law enforcement agencies share their data so MnDOT can compare the WIM record with the static weight. This process can tell MnDOT when WIM calibration is becoming an issue. MnDOT has found that truck drivers learn to avoid WIM sites when law enforcement is present. It does not take long for the word to spread. Truckers know the times of day that law enforcement may be present and avoid those times. There are several fracking sand mines in one area and the mine installed a truck scale so they would not be caught running overweight. Bad publicity lowered the number of overweight trucks at least for a time. There is a civil penalty for overweight loads and many truckers go to court, but many citations are dismissed. Pavement design still uses ESALs, and WIM data are used to develop the ESAL factors, but MnDOT is not currently using WIM load spectrum for design. MnDOT did one study using WIM data in concrete pavement design, but using WIM did not seem to affect the design substantially. MnDOT Bridge Office staff are interested in instances of very heavy loads, but not WIM for routine bridge design. MnDOT would like to begin using WIM in freight planning in the future. MnDOT systems with cameras can be used to identify some commodities when either logos or specific types of trucks are captured (milk, branded vehicles, logging trucks). MnDOT has used some photos to see that the system was misclassifying some vehicles as Class 8 trucks when photos showed they were dually pickup trucks pulling a boat or horse trailer. MnDOT used this exercise to tweak their classification algorithm. Florida The Florida DOT (FDOT) was selected for a follow-up interview based on meeting the criteria of proximity to East and Gulf Coast ports, and moving significant agricultural freight flows. The interview and survey produced the following case example. Sensors and Systems FDOT has one bending plate system still installed (not operational), and will remove it during the next roadway project involving that WIM location. All other WIM locations use quartz piezo-electric sensors. FDOT uses one specific brand because of proven performance. Other quartz piezo-electric sensors are being evaluated on an ongoing basis at an FDOT field test site, as well as another sensor type, which has pavement temperature sensitivities that affect the ability to obtain consistent weight accuracy. FDOT has brought in a consultant to perform QC on the LTPP WIM data, and to develop a standardized QC methodology to move forward with a new software application for traffic data polling and analysis. FDOT had a problem with updating the legacy software used to acquire the data from the WIM and other continuous count sites.
Case Examples 31 FDOT has been able to acquire federal highway project money for site installation projects, but money for maintenance has been a problem. FDOT has used state funds for the ongoing maintenance needs of the continuous count system. FDOT is moving from data transfer via a cellular modem system to fiber optic cable where available. This buried system is more reliable in emergencies like hurricanes, floods, wind- storms, and fires. It can be used for video transmission as well. Fiber is only available along major corridors, and has a significant connection cost as well, so it will not replace cellular service entirely. System calibration has been a problem, but FDOT hired a consultant to prototype calibration procedures for them to ensure FDOT was doing it correctly. While FDOT used to contract a calibration truck, the agency now uses the new procedures and their own people, including an FDOT truck from their Structures Group. With the new procedures and an FDOT truck and staff, FDOT can complete a calibration in half the time at less than half the cost. FDOTâs goal is to stay on a one-year calibration schedule. FDOT is funding a study on a method of continuous calibration, which uses Bluetooth identification technology to identify specific trucks in order to compare their weight going through enforcement weigh stations against their weight when going over WIM sites. FDOT can potentially eliminate the use of calibration trucks with a data pool of hundreds of vehicles each day. One thing FDOT has found is that the enforcement scales and WIM should be some- what close together. The farther apart they are, the more fuel consumption must be accounted for between the vehicle weights. FDOT reported that roadway deterioration is a problem in maintaining WIM systems and good data. FDOT studies the output across time in the WIM QC process. If FDOT sees differences between axles that do not make sense, it may indicate that the vehicle is bouncing at different speeds and a pavement that is not smooth. If FDOT sees this before a calibration, it will do a pre-site pavement inspection; if the pavement is not smooth, FDOT will grind it. FDOT believes north Florida is covered well in their current WIM network, but the networks in south and central Florida need expansion. The agency hopes to capture the large flower import market flows emanating from Miami, for example. Data Usage FDOT indicated that where and when they can supply good data to their internal customers, those customers want more. FDOT has begun using the AASHTO Pavement ME Design software, which allows for input of an axle load spectrum for pavement design (Cunagin et al., 2013). One developing concept at FDOT is using WIM data to study freight flows. FDOT has a research project with the University of Florida (UF) using artificial intelligence (AI) to look at WIM and camera image data. UF can spot differences and commodities in several ways, including axle weights, truck types, and logos. UF has noted differences among different types of haulers. Liquid haulers, for example, often show commodity logos or symbols, and the trucks can be quite specific in shape/size. The study right now has reviewed the capture of images, and UF is developing a database of vehicle types. UF can use WIM data to estimate partial loads, full loads, and cubed-out loads by the individual axle weights and differences in the axle weights on a given truck.
32 Use of Weigh-in-Motion Data for Pavement, Bridge, Weight Enforcement, and Freight Logistics Applications Maryland The Maryland DOT was selected for a follow-up interview based on several of the selection criteria. Maryland is a relatively small East Coast state with a port, significant freight traffic that is highly urbanized, and a developed WIM program. The combination of survey and interview produced the following status on the agencyâs use of WIM. Sensors and Systems In the survey, Maryland DOT had indicated that they used only quartz piezo-electric WIM sensors. Initially, Maryland DOT reviewed types of sensors and saw that quartz piezo-electric sensors have a good success rate in Europe and Asia. They are relatively insensitive to tempera- ture changes and work well at speeds of 30 to 70 mph. Maryland DOT has found they work well in both asphalt and concrete pavements. Additionally, they can be installed in 4 hours and open to traffic in 8 hours. This is especially advantageous in the Maryland traffic conditions and environment. Load cells and other sensors generally require more installation time, so Maryland DOT does not use these other sensors. Maryland DOT designs their WIM installations with two sets of half-lane-width sensors (four total half-lane-width sensors) per lane in a staggered configuration. This allows for continued data collection even when one sensor fails. FHWA calls this a double threshold, staggered layout (FHWA, 2018b). Maryland has VWS systems and an array of ATR installations. Maryland DOTâs VWS is a WIM augmented by cameras, with infrared lighting for night operation, and a laser for over- height detection. The current camera data are not sufficient to identify cargo or capture license plate or DOT numbers. Maryland DOT has a large ATR system. All these data are available for use by the DOT for pavement design, asset management, and freight planning. Installations are competitively bid, and the successful bidders have been integrators that use one brand of quartz piezo-electric sensors. Maryland DOT also has long-term mainte- nance contracts for their installations that require replacement of field-replaceable elements (essentially everything but sensors) within 24 hours of notification. Calibration every 6 months, using Class 9 trucks of known weight, is also included in the contract. This calibration schedule considers the temperature dependence of sensors, allowing for warm- and cool-weather adjust- ments. The contracts are paid on a monthly rate basis. Data Usage All data are livestreamed to the University of Maryland Center for Advanced Transportation Technology (CATT) for hosting, storage, and data analytics. The CATT lab allows data access to interested users. Two primary users are the Maryland State Police and the Maryland Transportation Authority, which operates the toll roads in Maryland. These entities use this information for commercial motor vehicle weight enforcement operations. Maryland DOT gets feedback from these weight enforcement agencies on the agreement of WIM and static scales used for actual overweight ticketing. The WIM functions as a flag for possible action and the law enforcement uses static or portable scales to verify and ticket. The Maryland State Police and the Maryland Transpor- tation Authority tell the DOT when the WIM does not agree with the static or portable scales so the DOT can check WIM sensor calibration. The first action is to adjust the sensor output
Case Examples 33 span settings to see if this solves the problem; if not, then a recalibration with trucks of known weight is required. Maryland DOT on occasion will alert law enforcement to suspicious activity the DOT sees at WIM sites. One example is suspected overweight vehicles hitting the WIM sensors at off hours when truck drivers believe no one will be staffing weight enforcement sites. Maryland DOT also sees instances of unbalanced loads. This is where the left and right wheel loads are significantly different. These can be trucks that are purposely changing lanes over the sensors to produce false responses or can signal a truck load that is not secured and shifting in the truck. Unsecured loads can be a safety issue. Maryland DOT is in a transition from using the AASHTO 1993 Pavement Design Guide to the AASHTOWare Pavement ME Design software, but is not using it for pavement design yet. The 1993 Guide uses ESALs while the Pavement ME software can use a default traffic spectrum from LTPP sites or a traffic spectrum from a WIM site. Both ESALs and the traffic spectrum can be obtained from WIM. Maryland DOT is very satisfied with their network of traffic data collection. However, some internal users may only use data collected from ATRs and may not use WIM data. Tennessee Tennessee DOT (TDOT) was chosen for a follow-up interview because, in addition to meet- ing some of the other criteria, TDOT used to collect WIM data, then discontinued collecting data, and now are working to develop the TDOT WIM network once again. Sensors and Systems TDOT discontinued their WIM program in 2006. TDOT had used portable WIM systems whose data were not of sufficient quality to justify the time, money, and effort that the systems required. Because of this, WIM took a back seat to other traffic data collection program ele- ments. Few TDOT employees are left from 2006 in the area responsible for the WIM data col- lection to provide historical knowledge. TDOT currently collects traffic data from 103 continuous count stations. TDOT also collects road counts at more than 12,000 sites. This will be decreasing as the continuous count stations increase. In 2017, TDOT conducted an in-house review of traffic data collection and needs assess- ment that included WIM. TDOT reached out to different departments, especially pavement design. TDOT identified the two greatest needs for weight data as stemming from 1) MEPDG pavement design and 2) weight data as a possible HPMS requirement. TDOT wanted to be ready for this. Having the data would be better than not having the data. TDOT also framed their recommendation to restart WIM data collection by also using the data for freight plan- ning. The funding to resurrect the WIM program is coming from the freight area of TDOT. In the preliminary studies, information was needed for both freight and project design and TDOT was able to analyze data needs to justify how freight planning could cover the cost. TDOT performed a GIS study to determine the locations for the new WIM program, consid- ering the National Highway System freight network, ports of entry to the state, major urban areas, and high truck volume sites. In addition to the in-house review, TDOT has studied WIM systems and talked to vendors and other DOTs, and will potentially use quartz piezo-electric sensors due to durability, reliability, accuracy, and affordabilityâbut no final decision on technology has been made.
34 Use of Weigh-in-Motion Data for Pavement, Bridge, Weight Enforcement, and Freight Logistics Applications Data Usage In TDOTâs study to redevelop the WIM system, TDOT spoke to possible stakeholders that might be interested in the data from the proposed WIM network. Agency departments that provided input and had an interest were Long-Range Planning, Roadway Inventory (which is responsible for the traffic data collection and will be responsible for WIM), Freight Planning, Pavement Design, Maintenance, and Materials. The Bridge Design department did not see a need for the WIM data in their operations. TDOT will work with law enforcement agencies to locate WIM stations near existing static scales as a way to screen trucks. TDOT plans to share some of the WIM information with these agencies. TDOT is considering a pay-for-data contract to provide a turnkey system with maintenance and calibration included by the vendor or contractor. Case Examples Summary The case examples provided a broader picture of the WIM use for the selected agencies than the survey could. They allowed follow-up questions for a better understanding of an agencyâs position and lessons learned. Following are the insights garnered from the interviews: â¢ Sensor type choices are driven by sensor accuracy, system reliability, and lane closure time for installation and maintenance. Sensor accuracy and system reliability are not the same for all DOTs. Some sensors work better in certain applications, road types, and weather. In con- gested areas, DOTs cannot afford long lane closures and have developed orchestrated plans to minimize closure time. Using only one type of sensor enables an agency to standardize operations and optimize installation and maintenance time. â¢ WIM systems are costly and much of the information can be gained with ATRs for much less cost. However, if weight information is needed, ATRs do not measure weight, and only a WIM system will provide this information. â¢ There is more use of contracting for maintenance and calibration services for WIM systems and pay-for-data solutions being considered. This is due to funding, personnel, and dwin- dling expertise at DOTs. â¢ Three of the five agencies are working toward implementation of the AASHTO Pavement ME Design software (which supports the MEPDG) in some capacity, but the others are electing to continue using previous pavement design methods that use ESAL calculations. (WIM can be used to calculate ESALs for legacy pavement design methods and for traffic axle spectrum in the ME pavement design procedure.) â¢ According to either the interview and/or survey responses, only two of the five DOTs indi- cated that agency Bridge Design personnel have used WIM data for bridge design, and these are only in a few special high truck volume locations (not for standard bridges). â¢ Of the five agencies, only one agency is actively using WIM data for freight planning pur- poses, but others are interested in beginning to use some data in this way. Several agencies reported gaps in their network that need additional sites to provide better WIM coverage for this purpose. â¢ One agency indicated there may be opportunities to augment or even supplant WIM data with purchased freight and commodity data available from Big Data companies. â¢ FDOT is working on a plan to pair static truck weight systems with WIM systems for the same truck using Bluetooth signatures for identification; this approach may reduce or eliminate
Case Examples 35 the need to calibrate WIM stations. Others are sharing WIM and static scale data with law enforcement to check calibration, but this would likely only include trucks stopped for suspected overweight violation. â¢ FDOT is working with the University of Florida to use AI with WIM and image data to under- stand freight flows. This includes commodity type and load (empty, partial, and full trucks). This is an innovative use of the data that could benefit all DOTs. â¢ Caltrans is working on research with the University of California at Irvine to study whether more data can be collected using inductive loops to possibly study freight flows by truck type.