Authoring tools for immersive virtual, augmented and mixed reality: A systematic review of content creation platforms, emerging technologies, and their limitations
Ricardo Rodrigues, Guilherme Gonçalves, Rui Machado, Bruno Peixoto, Maximino Bessa, Miguel Melo · Computers & Graphics · 2026
Immersive Reality is increasingly adopted across diverse sectors, including industry, education, and training. However, creating immersive content is frequently identified as a complex task that requires specialized programming skills. This systematic review follows the PRISMA framework to survey the existing literature on authoring tools for immersive Virtual, Augmented, and Mixed Reality, critically analyzing their technical characteristics, integration of emerging technologies, and limitations across 135 eligible studies. The data reveals a strong emphasis on addressing these technical barriers, showing that 116 of the reviewed tools require no programming expertise and utilize visual or demonstrative interactions instead. Regarding interaction, controller-based input is the primary standard (83 studies) due to its precision for complex UI tasks, while hand tracking (41 studies) is also frequently employed to enable direct manipulation. Regarding emerging technologies, the review identifies 28 studies that utilize Artificial Intelligence (AI) to assist users, while digital twins (DTs) and WebXR systems are each featured in 11 studies. Furthermore, a limitation in expressiveness was identified in 84 studies, referring to the degree of functional and representational capacity that a tool supports. In these cases, tools lack sufficient flexibility to implement complex logic, advanced workflows, or custom behaviors, thereby constraining the range of tasks or use cases they can support. Finally, approximately 26% of the analyzed literature is characterized as preliminary studies, with the Training & Simulation sector reporting the highest proportion (40%). Ultimately, this review demonstrates that while most research has focused on making tools easy to use, these solutions still lack functional expressiveness and the ability to handle complex tasks, as well as empirical validation.