Central Asian Research Index

Scientific Journal of Astana IT University, 2026

INTELLIGENT DRONE ROUTE PLANNING IN URBAN ENVIRONMENTS: OPTIMIZATION AND SAFETY

Manara Seksembayeva, Zhaksylyk Amangaliyev, Miras Anuarbekov

Аннотация · Abstract

This study develops and evaluates an intelligent route-planning model for unmanned aerial vehicles operating in urban environments with the objective of minimizing total flight time while maintaining flight safety. The proposed approach addresses a practical last-mile delivery scenario in a smart-city context and is implemented using Google OR-Tools within a constrained Traveling Salesman Problem framework. Open geospatial data from OpenStreetMap are used to obtain building geometries, street topology, and available height-related attributes, while meteorological data from OpenWeatherMap are used to account for wind conditions in flight-time estimation. Safe cruising altitude is determined from the maximum surrounding obstacle height with an additional safety margin. If explicit building-height data are unavailable, height is approximated from the number of building levels. A prototype system was developed to visualize routes, estimate route distance, flight time, and delivery cost, and export missions in MAVLink/QGroundControl-compatible format. Experimental scenarios for real delivery points in the Astana metropolitan area, including Astana, Kosshy, and Koyandy demonstrate the practical feasibility of the proposed open-data-driven routing approach and show improvements over a baseline sequential routing strategy. The scientific contribution of the study lies in the integration of open geospatial building data, weather-aware flight-time estimation, and safety-oriented altitude selection into a reproducible urban UAV route-planning framework. The proposed method can support the development of safe and cost-efficient drone delivery services in dense urban environments. Experimental scenarios with increasing numbers of delivery points demonstrated that the proposed method remains computationally efficient, with sub-second solver runtime for the tested cases for practical smart-city logistics applications.

Сведения · Details