Integrating adaptive signal control and autonomous vehicles for urban congestion relief and carbon reduction

Technical Summary: Integrating Adaptive Signal Control and Autonomous Vehicles for Urban Congestion Relief and Carbon Reduction

Problem Statement
Urban transportation systems face a dual challenge: escalating traffic congestion and rising carbon emissions, exacerbated by rapid urbanization and motorization. While Adaptive Traffic Signal Control (ATSC) and Autonomous Vehicles (AVs) are recognized as promising solutions, their coordinated system-level impacts remain poorly understood. Existing research suffers from methodological fragmentation, often evaluating ATSC and AVs in isolation under simplified or exogenous assumptions. Furthermore, most studies are limited to microscopic simulations of single intersections or small networks, failing to capture city-scale emergent effects, cross-city heterogeneity, and the economic feasibility of large-scale deployment. There is a critical gap in understanding how the interaction between signal control and vehicle behavior reconfigures operating states to reduce stop-and-go dynamics, and how these benefits vary across diverse urban contexts.

Methodology
The authors developed a comprehensive city-scale evaluation framework integrating microscopic traffic simulation, vehicle-level emission estimation, and interpretable machine learning.

• Simulation Platform: The study utilized CBEngine, a microscopic traffic simulation platform, to model vehicle trajectories and interactions.
• Control Strategies: Five scenarios were constructed for a representative 17-km urban corridor in Shenzhen:
  1. Baseline: Fixed-time signal control (Webster method, updated every 15 minutes).
  2. WB: Periodically updated adaptive signal control.
  3. MP: Real-time adaptive signal control using the Max-Pressure algorithm, which dynamically adjusts phases based on upstream/downstream queue differences.
  4. AV: Autonomous vehicles with enhanced longitudinal control (higher acceleration/deceleration limits and faster response times via Adaptive Cruise Control logic) under fixed-time signals.
  5. MP_AV: Synergistic deployment combining real-time Max-Pressure control with AVs.
• Emission Modeling: Carbon emissions were estimated using the Motor Vehicle Emission Simulator (MOVES) framework, calculating Vehicle Specific Power (VSP) based on instantaneous speed and acceleration to derive emission rates.
• City-Scale Analysis: The framework was applied to 100 major cities in China. To analyze heterogeneity, the authors employed Light Gradient Boosting Machine (LightGBM) models with Shapley Additive exPlanations (SHAP) to interpret how urban features (e.g., road density, baseline speed, population density, GDP) influence performance metrics.
• Economic Evaluation: A cost-benefit analysis (BCA) was conducted, calculating public investment costs (signal system upgrades) against socio-economic benefits (travel time savings, carbon reduction, and fuel savings).

Key Results

• Synergistic Performance: In the Shenzhen corridor, the coordinated MP_AV strategy significantly outperformed single-technology approaches. It reduced average travel time by approximately 32% (from 28.93 min to 19.47 min) and per-vehicle CO₂ emissions by 36% (from 5,678 g to 3,621 g) compared to the Baseline.
• Mechanism of Improvement: The gains were not merely additive but resulted from a systemic reconfiguration of vehicle operating states. The coordinated strategy suppressed high-emission, high-delay regimes (such as prolonged idling and frequent acceleration-deceleration cycles) and enhanced traffic flow continuity. Specifically, total idling time dropped from 8.98 min to 1.88 min.
• Deployment Thresholds: A nonlinear relationship exists between the deployment ratio of adaptive control and system performance. System-wide speed improvements plateaued when adaptive control covered approximately 80% of key intersections during peak periods, suggesting that near-optimal performance can be achieved without full network coverage.
• Cross-City Heterogeneity: Across 100 Chinese cities, coordinated deployment yielded substantial but variable benefits. All cities saw travel speed improvements exceeding 20% and emission reductions greater than 10%. However, the magnitude of benefits depended heavily on urban characteristics:
  • Traffic Efficiency: Greatest in cities with lower baseline speeds and moderate road network density. Excessively dense networks or high proportions of four-leg intersections tended to diminish marginal gains.
  • Economic Viability: High Benefit-Cost Ratios (BCR) were observed in cities with moderate infrastructure supply and strong, but not saturated, economic value. The highest BCRs were concentrated in core cities of the Pearl River Delta and Yangtze River Delta.
• Economic Impact: Aggregated across the 100 cities, the strategy generates approximately 7 million hours of annual travel-time savings (valued at ~US$70 billion), alongside ~US$46.1 billion in environmental benefits and ~US$50.6 billion in fuel savings.

Significance and Claims
The paper claims that coordinated infrastructure-vehicle intelligence offers a viable pathway for large-scale congestion mitigation and decarbonization without relying on continued road expansion. The study demonstrates that the benefits arise from the joint restructuring of traffic control and vehicle operations, rather than the simple superposition of technologies.

Key contributions include:

1. System-Level Insight: Providing evidence that ATSC and AVs interact to suppress stop-and-go dynamics and reconfigure vehicle operating states, a mechanism often overlooked in decoupled evaluations.
2. Differentiated Deployment: Identifying that the efficacy of these technologies is highly context-dependent. The findings suggest that "one-size-fits-all" deployment is suboptimal; instead, strategies should prioritize cities with specific structural conditions (e.g., moderate road density, unsaturated traffic) and focus on high-impact intersections rather than full network coverage.
3. Policy Relevance: Offering robust, interpretable evidence to support the phased and differentiated deployment of intelligent transportation technologies, aligning with national goals for low-carbon urban transitions and smart-city governance.

The authors conclude that while large-scale deployment requires significant investment in infrastructure and coordination, the socio-economic returns—driven primarily by time savings but supported by emission and fuel reductions—justify the strategy in a wide range of urban contexts, particularly those with moderate infrastructure supply and high operational improvement potential.