Assembly affects the product's performance and reliability directly. The current assembly method based on geometric deviation quantity controlling, cannot guarantee the physical performance for complex aeronautical thin-walled structures effectively, such as assembly geometry accuracy and internal interactive stress. And assembly performance controlling is taken as the bottleneck problem that restricting the new aviation requirement of sub-millimeter assembly. In this paper, by proposing the accurate prediction and process-oriented adjustment&controlling strategies on assembly quality, construction on working mode with “quantifiable and controllable” characteristic was proposed, whose aim was to reduce the phenomenon of out-of-tolerance and deformation rebound error, and the ultimate goal is to reduce the uncertainty of assembly performance parameters. At the technical level, the academic development context and existing problems for assembling thin-walled structures were reviewed and analyzed, such as assembly process parameters optimization, assembly error transfer and accumulation, comprehensive adjustment on assembly quality, and virtual assembly simulation validation. Then the key future research trends for aeronautical structure assembly were also put forward, i.e. the force/deformation coordination among multi-type finite units for non-ideal assemblies, the dynamic construction of stiffness matrix for intermediate assemblies considering geometric nonlinearity, the adaptive balance on assembly performance driven by physical modeling and measured data, and the inverse optimization on assembly quality and parameters with intelligent data processing. This paper would lay a solid foundation for achieving the accurate assembly mode with the characteristics of “intelligent/scientific, and active/collaborative controlling on geometric shape and physical performance”, and higher assembly quality and efficiency could also be gained.