Yannick conducted an agent-based simulation using MATSim. The results presented here compare three scenarios for shifting car trips to bike sharing and evaluate their potential impacts on emissions reduction.
天美mv天美 researchers reveal how bike sharing could shape future emission reduction policies
While the electrification of passenger vehicles is widely regarded as a key strategy for decarbonizing urban mobility, it is not progressing at a sufficient pace. Alternative transportation, such as biking, offers an economical solution to expedite the transition. Read more...
Introduction & Background
Launched in 2023 with funding support from Mobi by Rogers, Mitacs, and NSERC, this project evaluates the emission-reduction potential of bike-sharing systems. Using modeling, we examined how much personal car travel can be shifted to bike share, how that shift translates into reductions in criteria air contaminants and greenhouse gas emissions, and how bike sharing compares, from an emissions perspective, with other strategies such as the electrification of personal vehicles. Three MASc students, Ali Azimi (emission reduction modeling), Yannick Ntibinsiga (agent-based traffic modeling), and Shedrach Ezenwali (life cycle analysis) led the research.
Project Findings
The research team has integrated extensive datasets from Mobi by Rogers, the City of Vancouver, and TransLink鈥檚 trip diary (2017) to build multiple models, including emission submodels, an agent-based traffic model, and a life cycle analysis (LCA) model. Here are our main findings:
- Vancouver鈥檚 bike-sharing system, despite a relatively modest fleet of about 2,600 bikes, has the potential to reduce greenhouse gas emissions by up to 4% of total urban transportation emissions within the areas served by the network.
- Depending on utilization levels, even without expanding the system, every 5 to 16 shared bikes can achieve the same CO2 reduction impact as removing one fossil-fuel vehicle from the road.
- Results from a calibrated agent-based MATSim model show that an integrated fare system, allowing a single payment for multimodal trips combining bike share and public transit, could shift about 1.1% of car trips to more sustainable modes.
- Life-cycle assessment shows that a large share of the system鈥檚 CO2 emissions comes from bike rebalancing with diesel trucks used to move bicycles between stations.
- Encouraging user-based self-rebalancing could reduce emissions and improve system efficiency.
- Our findings suggest that asking users to return bikes to preferred stations at distances greater than 500 metres can reduce CO2 emissions while also remaining economically viable by balancing user incentives against reduced truck activity.
Ali developed a machine-learning-based model using bike trip data to estimate the mode shift from car trips that are bikeable but not walkable to bike sharing. He then examined scenarios representing low system utilization (the current case), moderate growth, and high utilization without network expansion, and quantified the emissions-reduction potential for each case.
Shedrach conducted the life-cycle analysis and showed that bike rebalancing is the largest source of CO2 emissions across the entire bike-sharing system. He also quantified the potential to reduce these emissions through user-based self-rebalancing supported by incentives.
Knowledge Dessimination and publications
Theses
- Yannick Ntibansiga, January 2026, Mode Shift Through Integrated Micromobility Policies:An Activity-Based Agent Modelling Framework
- Seyed Ali Azimi, January 2026, Data-Driven Assessment of Shared Micromobility Systems: Evaluating Environmental Impacts and Spatio-Temporal Demand Patterns
Journal Papers
- Ndayiragije, Y. & Hosseini, V. (2026). Evaluating Emission Reduction Strategies Through Micro-Mobility Adoption: An Agent-Based Modeling Approach for Vancouver (In-preparation)
- Azimi, A., Ndayiragije, Y., Razzaghi, N., & Hosseini, V. (2026). An enhanced spatio-temporal model for emission reduction calculation of shared micromobility systems. (In-Review)
- Azimi, A., Shabani, A., & Hosseini, V. (2026). Event-Aware Spatio-Temporal Forecasting of Shared Micromobility Demand Using a Two-Stage Deep Learning Framework (In-preparation)
- Ezenwali, S., Razmi, A.R., Hosseini, V. (2026), Life cycle and cost efficiency assessment of bike sharing with user-centered rebalancing (In-review)
Conference papers
- Ndayiragije, Y., Ezenwali, S., Finck, E., & Hosseini, V. (2025). Solar-Powered Charging Stations for E-Bikes: A Case Study in the City of Vancouver, Canada. 60th Annual Conference of Canadian Transportation Research Forum
- Azimi, A., Ndayiragije, Y., Razzaghi, N., & Hosseini, V. (2025) Emission Reduction Estimation of Urban Bike-sharing Systems using Machine Learning (60th Annual Meeting of the Canadian Transportation Research Forum)
- Azimi, A., Shabani, A., & Hosseini, V., (2024) Prediction of bike-sharing demand at cluster level and distance travelled using machine learning (59th Annual Meeting of the Canadian Transportation Research Forum)
Partners:
Academic Collaborators:
