Promoting sustainable tourism by optimizing tour paths of UNESCO World heritage cultural sites in India

This study examines how the spatial organisation of India’s UNESCO World Heritage Cultural (WHC) sites can inform heritage circuit design, visitor dispersal, and destination management, with particular attention to urban heritage nodes. Because visiting all sites in a tour is impractical, the sites are partitioned into coherent clusters using K-means. Each cluster is formulated as an Asymmetric Travelling Salesman Problem (ATSP), solved with an elitist Genetic Algorithm (GA), and benchmarked against the Nearest Neighbour (NN) heuristic and Ant Colony Optimisation (ACO). The GA results are 0.4% (38 km distance reduction) and 3.76% (293.8 km distance reduction), respectively, and are improved compared to ACO and NN across all four clusters. The clusters and routes give a benchmark sequence of visits for each group of urban and non-urban heritage nodes. The model minimises distance alone and omits carrying capacity, transport mode, emissions, traffic, and visitor behaviour. Its planning implications are untested with visitor-flow data. Distance is a partial proxy for transport-related resource efficiency, not a measure of sustainability. City and destination managers can use the clusters and routes as an evidence base for city-level heritage circuits. They can also use them to pair high-profile and lesser-visited sites to support visitor dispersal, to stagger arrivals at high-pressure nodes, and to coordinate national, state, and city agencies. To our knowledge, this is the first application of a clustering-based GA framework to India’s UNESCO WHC network. It links visitor-movement research with route optimisation and translates routing outputs into planning questions for urban heritage tourism.