Simulation Study Implementing Transit Signal Priority on a Major Transit Corridor with Two Transit Lines on a Bus Only Lane in San Francisco Reduced Total Travel Time up to Nine Percent and Control Delay by 20 Percent.

Researchers Developed Three Strategies Using Signal Optimization Techniques to Facilitate the Movement of Buses, Pedestrians, and Bicyclists in Complete Streets.

Date Posted

Improved Analysis Methodologies and Strategies for Complete Street

Summary Information

Complete Streets feature safe, equitable travel options for all road users, from cyclists and pedestrians to buses and delivery vehicles. However, the available methodologies used to evaluate such implementations are not widely used because they often require extensive local data and are based on subjective indices that are not easily transferable. The Complete Streets implementation, however,  rely mostly on qualitative assessment and a limited evaluation based on performance measures. The objective of this study was to develop and test an improved methodology for evaluating the traffic performance of alternative designs for Complete Streets, and to develop signal control strategies to improve travel at signalized intersections for all users using both conventional and emerging technologies, including sensors fusing machine vision and radar data to provide road-user detection and classification. Historical data from 2006-2010 as well as data collected from sensors during 2020-2021 were used in this study.


The following three strategies were tested: 

  • Strategy 1, Arterial timing optimization without private vehicles; Strategy 2 - Transit signal priority (TSP) at signalized intersections, and Strategy 3 - Bicycle detection and priority.
  • In Strategy 1, San Francisco's Market Street, a major corridor with 14 signalized intersections that was converted to bus only route in 2020, was used. The bus only scenario was modeled in a signal optimization software for two cases, where average speeds of 10 and 25 mi/h were used to account for the lower bus speeds. San Francisco Metropolitan Transportation Authority (SFMTA)’s spreadsheet tool was used for delay estimation.
  • Strategy 2, applied to Geary Street, San Francisco, involved optimizing signal timings across 11 intersections over a 0.9-mile stretch with transit-only lanes. The optimized timings determined by the signal optimization software were input into a microscopic traffic simulation model for multiple Transit Signal Priority (TSP) scenarios.
  • In Strategy 3, the objective was to test and evaluate bicycle priority using the developed level of service measures by leveraging data collection efforts from a separate concurrent research project that was implementing emerging vehicle to infrastructure (V2I) communication and sensors that use machine vision and radar technologies.


Benefits presented below pertain to the first two strategies. The data collection for Strategy 3 experienced delays due to the COVID-19 pandemic and the findings, at the time of this report, remain inconclusive.

  • In Strategy 1, under bus-only Market Street case for the 25MPH scenario, the optimal cycle length was found to be 65 seconds, which reduced overall network intersection delay by approximately 35 percent. The eastbound Market Street approach delay was reduced by 65 percent and the westbound was reduced by 77 percent.
  • In Strategy 1, under bus-only network case for 10 mi/h, the optimal cycle length was again to be at 65 seconds, which reduced the traffic delay by approximately 34 percent, with an eastbound reduction of 71 percent and westbound reduction of 70 percent.
  • In comparison, when the SFMTA'S spreadsheet was used, it was found that the optimal 65 seconds cycle length reduced the overall intersection delay by approximately 27 percent. The eastbound Market Street delay was reduced by approximately 67 percent and the westbound was reduced by about 71 percent.
  • In Strategy 2, for Geary Street, a signal cycle length of 60 seconds for corridors with bus lanes and transit signal priority reduced delays by 20 percent. The stop delay for vehicles along the route was reduced by up to 52 percent. Although this signal cycle led to a slight increase in travel times for private vehicles, the overall level of service remained unaffected.
Goal Areas
Results Type
Deployment Locations