The Minnesota Department of Transportation (MnDOT) launched the Minnesota Bicycle and Pedestrian Counting Initiative (the Initiative) in 2011. The Initiative is a collaborative, state-wide effort to encourage and support non-motorized traffic monitoring by local, regional, and state governments and non-profit organizations. MnDOT has funded three projects to support the Initiative: 1. Methodologies for Counting Bicyclists and Pedestrians in Minnesota (2011-13) 2. Implementing Bicycle and Pedestrian Traffic Counts and Data Collection (2013-15) 3. Institutionalizing the Use of State and Local Pedestrian and Bicycle Traffic Counts (2014-16) This report summarizes the results of the second project, the Implementation study. The general goal of this project was to demonstrate the feasibility of using automated sensors to collect bicycle and pedestrian traffic data in Minnesota. The main objectives were to work with local units of government to: • Acquire and install sensors for automated counting of bicyclists and pedestrians • Calibrate and validate the sensors • Use portable sensors for short duration counts • Develop models for extrapolation of short duration counts • Integrate continuous count data into MnDOT traffic monitoring databases Key findings from this study include: • Automated sensors for monitoring bicycle and pedestrian traffic are available commercially at reasonable cost and can be deployed at both permanent and short-duration monitoring sites. The sensors deployed in Minnesota included inductive loop sensors for permanent monitoring of bicycle traffic on roads (Duluth, Eagan, Minneapolis); integrated passive infrared and inductive loop sensors for permanent monitoring of bicycle and pedestrian traffic on trails (Duluth, Rochester); pneumatic tube sensors for short-duration monitoring of bicycle traffic on roads or trails (Bemidji, Grand Marais, Hennepin County) and radio beam sensors for short-duration monitoring bicycle and pedestrian traffic on trails (Bemidji, Grand Marais, Rochester). Sensors varied in costs from $1,200 to $7,500, plus costs for installation. Installation costs were in the vicinity of $5,000 for inductive loops embedded in the road surface or trail service. Each of the permanent sensors installed during the project operated successfully year-round, although some experienced problems and required maintenance in the field. Vendors provided a range of support services, ranging from troubleshooting to automated analyses of data. Sensors that transmit data remotely offer potential to reduce labour costs for data collection and analysis but require payment of annual service fees. • Bicycle and pedestrian traffic volumes varied greatly across locations, with highest volumes observed on multiuse trails in urban areas (e.g., traffic volumes as high as several thousand individuals per day in Minneapolis and Duluth). Much lower volumes were observed on residential streets and county roads in smaller communities (e.g., fewer than 10 bicyclists per day at locations in Bemidji and Grand Marais). In Hennepin County, mean daily bicycle volumes at short-duration monitoring sites on roads ranged from less than 10 to more than 1,000, with weekend daily volumes at one location exceeding 1,600. • All sensors tested in the study produced reasonably accurate measures of bicycle and pedestrian traffic, although accuracy varied with the specific technology, care taken in deployment, maintenance following deployment, analytic methods used in analysis, and specifics of the configuration, including traffic volumes. Most technologies tended to undercount. Occlusion, or multiple users passing a sensor simultaneously, was a principal source of error. Correction equations and adjustment factors were developed to correct counts for systematic error, but the need for correction depends on the application, the need for accuracy, and the costs for additional data management. • Portable sensors were deployed efficiently and provided useful measures of bicycle and pedestrian traffic. Because of hourly and day-of-week variations in bicycle and pedestrian traffic, short-duration monitoring results are most useful if they include a minimum of seven complete days (i.e., the weekdays and weekend days). However, if the goal of monitoring is simply to obtain an indicator of the general order of magnitude of bicycle or pedestrian traffic, shorter monitoring periods may suffice. • The FHWA’s Traffic Monitoring Guide (2013) outlines an approach to estimating annual average daily traffic that involves use of factors derived from permanent monitoring locations to extrapolate short-duration counts. To demonstrate the applicability of this approach to non-motorized traffic monitoring, the study team used data from permanent and short-duration monitoring on an 80-mile multiuse trail network in Minneapolis to estimate Average Annual Daily Traffic (AADT) for each trail segment. These estimates then were used to estimate miles travelled by bicyclists and pedestrians on the network. The study team showed that AADT on trial segments ranged from about 40 to more than 3,700 and that summertime trail Annual Daily Traffic (ADT) was approximately double AADT. The study team estimated that users of the Minneapolis trail network travelled more than 28 million miles on the trails in 2013. The study team identified four traffic patterns on trails: commuter, commuter-mixed, multipurpose, and multipurpose-mixed. These patterns potentially can be used to establish factor groups for use in other applications. • A major challenge in implementing bicycle and pedestrian traffic monitoring is data management, specifically the challenge of formatting data from different sensors and integrating data into motorized traffic monitoring data management systems. The goal of integrating data collected during the Implementation study into MnDOT’s traffic data management was not achieved because the vendor that was supporting implementation of the new data management system ceased operations. In addition to these findings, an additional outcome from this project is a new MnDOT guidance document, “DRAFT Bicycle and Pedestrian Data Collection Manual.” Team members used this manual in training workshops in spring 2015. The DRAFT Manual includes a set of case studies that summarize how local officials have and are using bicycle and pedestrian accounts to inform transportation planning, engineering, and policy-making. These case studies illustrate the demand for better data that exists in Minnesota. Years will be required to implement and institutionalize bicycle and pedestrian traffic successfully. One strategy that may be useful as staff work to institutionalize monitoring is to complete case studies that illustrate how officials have used data to make decisions on projects that increase the efficiency and safety of transportation systems. (Author/publisher)
Abstract