Industry 4.0 technologies are key elements for companies’ competitiveness. Among these technologies, Augmented Reality (AR) already shows great potential and results to assist workers through a large panel of industrial processes. Latest research suggests that AR systems should meet the user’s needs to deliver a personalized experience and therefore improve the adoption of the technology. Nevertheless, there is still a lack of theoretical guidance to design AR-based assistance capable of delivering both the right level of details of information and the best information design, based on the user’s needs. To address this problematic, we developed a methodology describing the principle of adaptive assistance as well as standard characteristics and criteria to design Augmented Reality-based Adaptive Assistance Systems (ARAAS). We applied this methodology on an industrial maintenance use case in the context of landing gears’ overhaul operation. The feedbacks from different profiles of inspectors underline the usefulness of such system and its capacity to deliver an assistance adapted to their needs.