Abstract
The transition toward autonomous electric mobility fleets promises transformative reductions in urban congestion and tailpipe emissions, yet operationalizing these systems remains constrained by fragmented optimization paradigms. Existing frameworks typically isolate spatial fleet dispatch from vehicle-level hardware degradation and institutional capital governance. This study proposes a socio-technical synthesis integrating multi-agent deep reinforcement learning for dynamic rebalancing, edge-computing diagnostics for real-time component prognostics, and engineering-accounting governance mechanisms to safeguard capital efficiency. By examining the interplay between algorithmic dispatch, vehicular physical durability, spanning power electronics, thermal management, and structural glazing and financial oversight, this research highlights how uncoordinated technical optimization can inadvertently accelerate asset depreciation and fiscal overruns. While empirical validation across heterogeneous urban networks remains ongoing, this integrated perspective suggests that sustainable fleet deployment requires aligning decentralized computational intelligence with rigorous lifecycle cost governance and institutional adaptability.

This work is licensed under a Creative Commons Attribution 4.0 International License.
Copyright (c) 2026 Alex Mercer (Author)