<p>One of the major concerns for the IoT connectivity of the next-generation smart cities is how well the aerial networking could perform when the existing infrastructures are damaged in the scenario of a post-disaster. To this end, our work proposes a novel Define-Measure-Analyze-Improve-Control (DMAIC) enhanced framework for autonomous unmanned aerial vehicle (UAV) positioning that electrically directs the floating base station in the 3D space in the most coverage and energy-efficient manner. Essentially, the model simultaneously optimizes the UAV’s x-y coordinates and altitude by implementing analytical partial-derivative optimization over realistic urban channel conditions, such as elevation-dependent path loss and mixed Rician/Rayleigh fading. Extensive simulations reveal that the proposed methodology outperforms fixed-altitude deployment strategies in energy efficiency and coverage probability by as much as 55 % when the latter is used as a benchmark. Moreover, the DMAIC method also frequently outperforms a state-of-the-art UAV positioning baseline in terms of performance metrics: (5–10)% higher coverage, (10–15)% better normalized energy efficiency, up to 3 ms lower latency at peak loads, and around 7–10 Mbps higher aggregate throughput with increasing IoT density. Besides providing valuable design principles for the deployment of robust, energy-efficient UAV-assisted networks in 6G-enabled smart city infrastructures, which are indispensable for the scenarios of an emergency and a network recovery, the present findings serve as evidence for the effectiveness of analytically guided, adaptive UAV repositioning in complex urban environments.</p>