Equity- and Robustness-Aware Planning for Urban Medical Drone Services: Spatial Screening and Workflow Insights from Shenzhen

Early-stage planning for metropolitan medical-drone networks can use demographic and healthcare-facility data to identify equitable and robust candidate nodes while operational evidence is developing. This study presents an equity- and robustness-aware screening framework for candidate healthcare sites in Shenzhen using population, age structure, administrative boundaries, and medical point-of-interest data for 74 sub-districts. A station is defined as a planning-stage dispatch or handover candidate. The demand index is evaluated alongside total population, older population, and equity-balanced objectives. Across 96 configurations that vary candidate pool, service radius, station budget, and objective, we measure citywide and worst-district coverage, solution stability, minimax regret, redundant coverage, and single-node failure. At a 5 km radius and K = 15, equity-balanced screening increases worst-district population coverage from 11.0% to 52.3% in the core pool and from 24.1% to 70.5% in the mixed pool, while retaining 92.0% and 93.8% citywide coverage. The core minimax plan reduces maximum regret from 43.6 to 12.6 percentage points. Single-node analysis identifies concentrated failure exposure, while expansion of the core equity plan from K = 15 to K = 20 raises redundant population coverage from 1.0% to 13.5%, demonstrating a practical pathway for strengthening backup reach. A scenario-based workflow comparison further isolates the trade-off between nearest-station and load-aware dispatch: the latter improves assignment balance with a mean completion-time increase of 0.27 min when other capabilities are equalized. Together, the framework provides an actionable candidate shortlist, identifies priorities for reserve-node placement, and supports the design of subsequent site, route, and workflow pilots.