Road maintenance operators often work in dangerous environments and are in need of a support system to enhance their safety and efficiency. Augmented reality (AR) has proven to be useful in providing support to operators in various industrial sectors. However, the vast majority of the existing applications focus mainly on static, controlled environments, such as industrial shopfloors, although the dynamic flow of information that AR can provide could be very valuable to unstructured, dynamic environments. This paper presents a novel AR—based methodology for human–robot collaboration, real‐time instructions, and support for road maintenance operations, aiming to enhance operator safety and efficiency. A robot operating system‐based architecture is exploited for the communication of the modules. The methodology is tested in a laboratory environment, and the results validate the hypothesized enhancement of road operators. In the future, the application will be validated on real interventions in highways.