Technological innovation capability (TIC) is critical for decarbonising transport systems and advancing regional sustainability in China. However, existing research primarily focuses on the isolated direct impacts of TIC on transport carbon emissions (TCE), neglecting complex spatial interdependencies and typically decoupling the production and consumption sides. To bridge this research gap, this study develops a novel integrated production-consumption analytical framework to elucidate how TIC generates synergistic emission reduction benefits through spatial interactions. Using panel data from 30 provincial-level administrative regions in China, this research introduces methodological innovations by combining social network analysis to map inter-provincial emission topologies with a Spatial Durbin Model (SDM) evaluated across multiple weight matrices. This approach captures the direct effects, spatial spillovers, and interaction effects of TIC alongside key production-side (transport structure, energy efficiency) and consumption-side (industrial structure, household consumption) factors. The results indicate that China's provincial TCE display significant strengthening of spatial agglomeration based on the adjacency weight matrix. While TIC increases local TCE, it notably generates negative spatial spillovers. Crucially, production-side interactions extend carbon reduction benefits to surrounding provinces via a "hard technology" diffusion mechanism. By contrast, consumption-side interactions, constrained by rebound effects and high-frequency logistics demand, tend to offset TIC's mitigation potential and may even induce positive emission spillovers. These findings conclude that achieving carbon-neutral transport requires transitioning from isolated provincial governance to cross-regional collaborative strategies. Differentiated, regionally integrated policies that align "hard" production-side innovation incentives with "soft" consumption-side behavioral interventions are essential to unlock the spatial spillover potential of TIC.
