Auction-Driven ADMM Framework for Plan Consensus in Integrated Ground–Air Mobility

Abstract This study introduces a dispatcher-mediated plan consensus strategy for coordinating vehicle planning in autonomous mobility-on-demand systems. The approach employs the Alternating Direction Method of Multipliers (ADMM) to enable efficient plan consensus within integrated ground–air mobility networks. The proposed method combines auction-driven consensus protocols with ADMM-based decomposition techniques, applying ADMM directly to discrete plan variables. The framework integrates decomposition-based optimization with market-based plan consensus coordination in a setting where vehicles exchange only plan-level allocations and price signals, without sharing detailed contextual information, while the dispatcher coordinates request-level consistency through the consensus update. By incorporating auction-based price signals into this update, the dispatcher can account for diverse vehicle-local plans and guide request allocation in a way that reflects both local feasibility and system-level objectives. Simulation results show that the method can generate cooperative dispatch solutions with significantly reduced centralized computation in the tested scenarios, supporting its use as a dispatcher-mediated coordination mechanism for integrated ground–air mobility operations.