In this modern era, the constant increase in computational and sensing power has lead to the development of multiple data-driven concepts. Amongst these, the ‘digital thread’ is an architecture that aims at optimising the knowledge of a system by merging prior knowledge of the product with information from multiple stages of its lifecycle in order to improve the performance of new products to be designed, thanks to the increased accuracy of this updated knowledge. Even though the use of these data-driven architectures is becoming increasingly widespread, most of the corresponding developments remain currently limited to the component level. In this respect, this article extends the application of the digital thread from the component level to the structural assembly level and enriches it with additional multi-physics considerations and non-linear failure constraints. To this end, it details the development of a digital thread dedicated to an aircraft vertical tail plane structure made of carbon fibre reinforced polymer, as well as the tools and models required to implement such an approach using Bayesian inference, multi-physics simulations, and empirical models.