Reducing CO2 emissions from the rebalancing operation of the bike-sharing system in Beijing

Development of bike-sharing systems (BSS) and emphasis on green, low-carbon technology is increasing.
Environmental impacts of BSS have garnered attention.
Rebalancing operations often use fossil-fuel vehicles, leading to emissions overlooked by current studies.
Previous analysis of Bike-sharing Rebalancing Problem (BRP) focused on algorithms neglecting real-world applicability of models.
New method proposed to optimize CO2 emissions from BRP, tested in Beijing's BSS.
- Key findings:
  - High potential for optimization in Beijing’s BSS due to poor planning.
  - CO2 emissions could be reduced by 57.5% through improved placement of parking nodes and reduced vehicle routes.
  - Specific districts require increased bike launch quantities.

China became the largest carbon emitter by 2006 with emissions reaching 13.92 billion tons in 2019, slightly decreasing during COVID-19.
Transportation is a major CO2 emitter.
Beijing's ring-shaped roads lead to traffic congestion, negatively impacting air quality.
Traditional traffic management approaches in Beijing have shown limited results.
The sharing economy, including BSS, has emerged as a viable solution to transport and emissions issues.

BSS is gaining popularity as a solution to urban mobility and environmental challenges.
China leads globally in the number of bicycles with over 400 BSSs.
The BSS addresses the 'last mile' transportation challenge, promoting environmentally friendly travel.
The BRP seeks to balance bike supply and demand at parking nodes, remains an NP-hard problem.

Limited focus on environmental impacts of the rebalancing operations of BSS.
Studies used small, generated datasets and applied numerical analysis without real-world data, limiting relevance.
Existing models assume fixed stations and are ill-suited to dockless systems.
Traditional models overlook the trade-off between unmet demand and emissions.

Proposes to use real BSS data from Mobike in Beijing to address these gaps.
Aims to analyze temporal and spatial characteristics of BSS and optimize CO2 emissions while meeting demand.
Introduces a partitioning strategy to reduce problem complexity in BRP.
Aims to develop practical contributions applicable to other urban settings.

Prior studies emphasize direct impact but ignore indirect consequences such as increased motor vehicle usage for rebalancing.
Research shows rebalancing operations could negate benefits from shared bike rides.
Most studies lacked real-world data.

Static rebalancing occurs at night, while dynamic occurs during busy hours.
Static is more manageable and less polluting.
Research has primarily focused on static rebalancing, neglecting dynamic aspects.

Current models assume fixed vehicle routes and only minimize demand within predetermined constraints.
Real-world conditions in dockless systems alter these assumptions significantly.

Standard algorithms are not feasible for large-scale applications; heuristic methods are more practical.
Tabu search methods noted for effectiveness in solving routes.

Real order data from Mobike used for analysis, covering significant transactions and user details during May 2017.

Utilizes Manhattan distance calculations for estimating travel distances within Beijing’s urban layout.

Trip start times peaked during commuting hours, differing significantly on weekends.
Spatial distribution indicates higher bike demand in central districts.

Current and optimization methods proposed, focusing on minimizing emissions while meeting bike demands.
A partitioning strategy aims to facilitate operational efficiency.

Findings demonstrate a critical need for optimized rebalancing in BSS to achieve emission reductions.
Rebalancing operations may need to be refined for individual districts based on trip demand.
Introduced methods provide practical applications for city planners aiming to enhance bike-sharing systems.