Frank Proulx
Research Subjects
Frank Proulx has a total of 17 co-authors. Here are the top co-authors by number of shared publications:
- Mintu Miah
Univ. of California
2 shared publications - Offer Grembek
University of California Berkeley
2 shared publications Ipsita Banerjee
Univ. of California, Berkeley
1 shared publicationJohn Bigham
Univ. of California, Berkeley
1 shared publication- Joseph Broach
Portland State Univ
1 shared publication - Julia B. Griswold
Safe Transportation Research and Education Center, University of California, Berkeley, CA
1 shared publication
Latest Publications
- 1 November 2025
- Published by American Society of Civil Engineers (ASCE) in Journal of Transportation Engineering, Part A: Systems
- Vol. 151 (11)
Methodology of Large-Scale Annual Average Daily Bicycle Traffic Estimation
- 1 November 2025
- Published by American Society of Civil Engineers (ASCE) in Journal of Transportation Engineering, Part A: Systems
- Vol. 151 (11)
Methodology of Large-Scale Annual Average Daily Bicycle Traffic Estimation
- 10 July 2023
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2678 (3) , 573-589
Evaluating the Potential of Crowdsourced Data to Estimate Network-Wide Bicycle Volumes
This research integrated and evaluated emerging user data sources (Strava Metro, StreetLight, and hybrid docked/dockless bike share) of bicycle activity data with conventional “static” demand determinants (land use, built environment, sociodemographics) and measures (permanent and short-duration counts) to estimate annual average daily bicycle traffic (AADBT). We selected six locations (Boulder, Charlotte, Dallas, Portland, Bend, and Eugene) covering varied urban and suburban contexts and specified three sets of Poisson regression models: city-specific models, an Oregon pooled model, and all cities pooled. Static variables, Strava, and StreetLight complemented one another, with each additional data source tending to improve the model performance. Sites with lower volumes were more difficult to predict, with considerable error in even the best-performing models. City-specific models in general exhibited improved fit and prediction performance. Expected prediction error increased by a factor of about 1.4 when using Strava or StreetLight alone, but without static adjustment variables, to predict AADBT. Combining Strava plus StreetLight, but without static variables, increased error slightly less; by 1.3 times. We also found that transferring the model specifications from one year to the next without re-estimating the model parameters resulted in a 10% to 50% increase in error rates across models, so such transfer is not recommended. The findings from this study indicate that rather than replacing conventional bike data sources and count programs, old “small” data sources will likely be very important for big data sources like Strava and StreetLight to achieve their potential for predicting AADBT.
- 10 July 2023
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2678 (3) , 573-589
Evaluating the Potential of Crowdsourced Data to Estimate Network-Wide Bicycle Volumes
This research integrated and evaluated emerging user data sources (Strava Metro, StreetLight, and hybrid docked/dockless bike share) of bicycle activity data with conventional “static” demand determinants (land use, built environment, sociodemographics) and measures (permanent and short-duration counts) to estimate annual average daily bicycle traffic (AADBT). We selected six locations (Boulder, Charlotte, Dallas, Portland, Bend, and Eugene) covering varied urban and suburban contexts and specified three sets of Poisson regression models: city-specific models, an Oregon pooled model, and all cities pooled. Static variables, Strava, and StreetLight complemented one another, with each additional data source tending to improve the model performance. Sites with lower volumes were more difficult to predict, with considerable error in even the best-performing models. City-specific models in general exhibited improved fit and prediction performance. Expected prediction error increased by a factor of about 1.4 when using Strava or StreetLight alone, but without static adjustment variables, to predict AADBT. Combining Strava plus StreetLight, but without static variables, increased error slightly less; by 1.3 times. We also found that transferring the model specifications from one year to the next without re-estimating the model parameters resulted in a 10% to 50% increase in error rates across models, so such transfer is not recommended. The findings from this study indicate that rather than replacing conventional bike data sources and count programs, old “small” data sources will likely be very important for big data sources like Strava and StreetLight to achieve their potential for predicting AADBT.
- 22 February 2022
- Published by Elsevier in Transport Policy
- Vol. 120, 23-39
Pedestrian fatalities in darkness: What do we know, and what can be done?
- 22 February 2022
- Published by Elsevier in Transport Policy
- Vol. 120, 23-39
Pedestrian fatalities in darkness: What do we know, and what can be done?
- Open Access
- 1 January 2015
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2527 (1) , 99-106
Database for Active Transportation Infrastructure and Volume
Information about pedestrian infrastructure and volume is indispensable to monitoring, evaluating, and improving the environment for comfortable and safe walking. However, determining and organizing the various types of data in a way that is easy to update and analyze can present challenges. This study designed and developed a relational database for pedestrian infrastructure and volume and included two core components (node table and approach table) and several subcomponents (tables for crosswalks, sidewalks, buffers, signs, transits, bikeways, bicycle parking, and volumes). Important measurements were proposed on the basis of the literature and practice review and grouped into different component categories on the basis of their attributes and relationships. Links were defined according to their relative locations to connect all the components. An infrastructure data collection pilot was conducted across 100 mi (161 km) of California highways using computer imagery, and across 7 mi (11.27 km) of those highways by field inventory, to prove the feasibility of the database. Time costs associated with collecting infrastructure data for the entire state highway system were estimated to be 4,006 h and 8,935 h for using computer and field collection methods, respectively. This study demonstrated that the database was easy to maintain, flexible to update, and feasible for data collection both by computer imagery and in the field. Although most data in the database were related to pedestrians, basic bicyclist-related information was also included to demonstrate the transferability of the database to store bicyclist infrastructure and volume in the future.
- 1 January 2015
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2527 (1) , 99-106
Open AccessDatabase for Active Transportation Infrastructure and Volume
Information about pedestrian infrastructure and volume is indispensable to monitoring, evaluating, and improving the environment for comfortable and safe walking. However, determining and organizing the various types of data in a way that is easy to update and analyze can present challenges. This study designed and developed a relational database for pedestrian infrastructure and volume and included two core components (node table and approach table) and several subcomponents (tables for crosswalks, sidewalks, buffers, signs, transits, bikeways, bicycle parking, and volumes). Important measurements were proposed on the basis of the literature and practice review and grouped into different component categories on the basis of their attributes and relationships. Links were defined according to their relative locations to connect all the components. An infrastructure data collection pilot was conducted across 100 mi (161 km) of California highways using computer imagery, and across 7 mi (11.27 km) of those highways by field inventory, to prove the feasibility of the database. Time costs associated with collecting infrastructure data for the entire state highway system were estimated to be 4,006 h and 8,935 h for using computer and field collection methods, respectively. This study demonstrated that the database was easy to maintain, flexible to update, and feasible for data collection both by computer imagery and in the field. Although most data in the database were related to pedestrians, basic bicyclist-related information was also included to demonstrate the transferability of the database to store bicyclist infrastructure and volume in the future.
- 1 January 2013
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2339 (1) , 68-79
Classification of Bicycle Traffic Patterns in Five North American Cities
This study used a unique database of long-term bicycle counts from 38 locations in five North American cities and along the Route Verte in Quebec, Canada, to analyze bicycle ridership patterns. The cities in the study were Montreal, Quebec; Ottawa, Ontario; and Vancouver, British Columbia, in Canada and Portland, Oregon, and San Francisco, California, in the United States. Count data showed that the bicycle volume patterns at each location could be classified as utilitarian, mixed utilitarian, mixed recreational, and recreational. Study locations classified by these categories were found to have consistent hourly and weekly traffic patterns across cities, despite considerable differences between the cities in their weather, size, and urban form. Seasonal patterns across the four categories and in the cities also were identified. Expansion factors for each classification are presented by hour and day of the week. Monthly expansion factors are presented for each city. Finally, traffic volume characteristics are presented for comparison purposes.
- 1 January 2013
- Published by SAGE Publications in Transportation Research Record: Journal of the Transportation Research Board
- Vol. 2339 (1) , 68-79
Classification of Bicycle Traffic Patterns in Five North American Cities
This study used a unique database of long-term bicycle counts from 38 locations in five North American cities and along the Route Verte in Quebec, Canada, to analyze bicycle ridership patterns. The cities in the study were Montreal, Quebec; Ottawa, Ontario; and Vancouver, British Columbia, in Canada and Portland, Oregon, and San Francisco, California, in the United States. Count data showed that the bicycle volume patterns at each location could be classified as utilitarian, mixed utilitarian, mixed recreational, and recreational. Study locations classified by these categories were found to have consistent hourly and weekly traffic patterns across cities, despite considerable differences between the cities in their weather, size, and urban form. Seasonal patterns across the four categories and in the cities also were identified. Expansion factors for each classification are presented by hour and day of the week. Monthly expansion factors are presented for each city. Finally, traffic volume characteristics are presented for comparison purposes.