Carbon emission-focused dynamic eco-route guidance design for urban expressway systems

Eco-routing offers a promising approach to reducing carbon emissions. However, most routing systems theoretically overlook the potential conflict between emission reduction and traffic efficiency, leading to either ‘eco-friendly but slow’ or ‘fast but eco-unfriendly’ outcomes; In practice, existing systems generally rely on real-time or historical data, potentially causing ‘navigation-induced emission/congestion’ as many vehicles are directed toward temporarily ‘good route’. To address this, this paper introduces a multi-agent Markov decision process (MMDP) for dynamically optimising system-wide route guidance (RG) strategies. The framework models both eco-RG and delay-RG for urban expressway systems and reduces congestion by providing differentiated RG information. A case study demonstrates that the proposed model effectively balances traffic flow, improves efficiency, and reduces emissions. The eco-RG strategy achieves the greatest emission reductions while maintaining efficiency. Notably, the study reveals a trade-off between eco-RG and delay-RG and compares the proposed method with app-based RG systems, offering insights for eco-friendly navigation.