Evolutionary multicriteria planning of bus stops location in smart cities

In modern smart cities, public transportation systems are the foundation of mobility services. Specifically, public bus services are crucial for improving accessibility and the quality of life for many residents, as they provide an affordable and flexible means of travel. A primary consideration in bus network planning involves strategically determining the optimal placement of bus stops throughout the service area. Conventional bus network planning prioritized system operator cost-efficiency as the primary objective, relegating user quality of service to a subordinate goal. Conversely, recent modeling efforts have shifted focus toward user-centric objectives to facilitate the principles of the Transit Oriented Development paradigm. This article introduces a solution to the multicriteria Bus Stop Location Problem, to find the best placement for bus stops to effectively balance user accessibility, service efficiency, and operational costs. To solve this complex trade-off, a multi-objective evolutionary algorithm is developed. The algorithm simultaneously optimizes three key, often conflicting, objectives: i) maximize demand coverage, to ensure the service reaches the most potential users, ii) minimize travel times, by reducing the total number of stops, and iii) minimize operational costs, measured and weighted based on passenger demand. The research applies the resolution approach to real-world case studies using actual public transport mobility data from Montevideo, Uruguay, and Buenos Aires, Argentina. The experimental results confirm that the proposed multi-objective evolutionary algorithm is effective, reliably generating high-quality Pareto fronts that capture the significant trade-offs among the objectives: demand coverage, travel time, and operational cost. Compared with a local search heuristic and a single-objective evolutionary algorithm, superior convergence and diversity were obtained. The compromise solution improved up to 46.9% in the coverage-to-cost ratio compared with the local search baseline, highlighting the advantage of the multi-objective approach for balanced and cost-efficient planning decisions. The computed solutions exhibit a tendency to prioritize services in high-demand central urban areas while simultaneously offering policymakers a diverse set of feasible design alternatives.