The long-term performance of pavement structures is influenced by the combined effects of repeated traffic loads and environmental conditions, such as temperature variations. Traditional numerical approaches face significant challenges in simulating the decades-long lifespan of pavements due to the vast disparity between the time scales of individual loading events and the overall lifespan. This paper presents a novel framework for the thermo-mechanical analysis of pavement structures, based on a time homogenization technique to efficiently predict long-term performance. An Arbitrary Lagrangian-Eulerian (ALE) formulation is employed to simulate the mechanical response of pavements to consecutive tire overruns, while a thermo-mechanical material model accounts for the effects of temperature variations. A key innovation is the treatment of history variables (HVs). The framework incorporates both spatial and temporal mapping of HVs, enabling the extrapolation of material behavior when the evolution rate of HVs stabilizes. This allows for the skipping of computationally expensive tire overruns, significantly reducing simulation time without compromising accuracy. Numerical examples demonstrate the framework's ability to predict the long-term response of pavement structures under realistic traffic and thermal conditions.