Cross Reality Seeks a Foothold on the Factory Floor, but the Evidence Is Still Thin

Imagine an engineer who begins her morning inside a fully immersive virtual factory, rearranging production lines with a flick of a controller, then slips seamlessly into an augmented reality view of the real shop floor to check whether a new machine actually fits the available space. That fluid movement between degrees of virtuality is the promise of Cross Reality, or CR, an emerging branch of immersive computing that lets users transition between augmented reality (AR) and virtual reality (VR) without breaking their workflow. A new open-access study published in the Journal of Ambient Intelligence and Humanized Computing by researchers at the Polytechnic University of Bari asks a deceptively simple question: is cross reality actually crossing into industry? The answer, based on a systematic review of the literature and a structured design workshop with experts, is a cautious not yet.

The research team, led by Luana Marangelli together with Enricoandrea Laviola, Antonio Emmanuele Uva and Michele Gattullo, framed their investigation as a conceptual design study grounded in a Systematic Literature Review following the PRISMA framework. They searched five major databases, including Scopus, IEEE Xplore and Web of Science, in December 2025, using a search string that paired the explicit term Cross Reality with industrial keywords such as assembly, maintenance, training, logistics and prototyping. From an initial pool of 402 records, refined to 281 unique papers, only 16 studies published between 2009 and 2025 survived screening and full-text assessment. A complementary search of grey literature, including company websites and blog posts, yielded no concrete examples of CR deployed in real industrial settings at all.

That small number is the study’s most striking finding. Despite a surge of interest in immersive technologies across manufacturing, logistics and design, genuinely industrial applications of CR remain scarce and immature. The authors note that the term itself has shifted meaning over time: early work from 2009 and 2010, such as Paradiso and Landay’s cross-reality environments, emphasized human-machine collaboration rather than transitions between immersive states. Publication activity spiked around 2009 and 2010, then declined sharply until roughly 2019, when renewed attention emerged alongside modern head-mounted displays. The arrival of video see-through devices with passthrough capability, such as the Varjo headsets, Apple Vision Pro and Oculus Quest, has accelerated the trend by allowing a single headset to render both AR overlays and fully immersive VR scenes.

Technologically, the review traces a clear evolution in enabling hardware. Before 2018, industrial CR-like systems relied on mobile phones, projectors and tablets. Optical see-through devices such as the Microsoft HoloLens supported AR but could not deliver convincing VR. More recent optical see-through hardware like the MagicLeap 2 introduced a dimming mode that tints the display behind digital content, reducing the visibility of the real world and enhancing immersion, while video see-through headsets from Varjo and Meta now support genuine passthrough transitions. The 16 reviewed studies span procedural tasks, prototyping, authoring, process simulation, training and quality inspection, and collaboration emerged as a defining feature: nine of the 16 studies involved users interacting across physical and virtual environments.

The evidence base, however, remains exploratory. Only seven of the reviewed works reported experiments with end users, and just two fully satisfied the authors’ evaluation-rigor criterion, which considered sample adequacy, evaluation measures and documentation. Common metrics included the System Usability Scale, task completion time and the NASA Task Load Index. Some results are encouraging: a CR wire-assembly system evaluated by Simoes and colleagues with workers with disabilities produced measurable improvements in task-execution efficiency, and recent work by Wu and colleagues on XR-assisted remote quality inspection showed significantly reduced error rates and completion times compared with unassisted inspection, alongside improved social presence. But small samples and controlled settings limit generalizability, and only five studies were evaluated in actual industrial environments.

To move beyond the thin literature, the researchers ran a collaborative design session with 12 participants, five academics and seven industry professionals, each with at least five years of experience applying AR or VR in industrial settings. Across three meetings, the group first mapped the challenges and opportunities of AR and VR separately, then brainstormed prospective CR applications for seven industrial domains: authoring, discrepancy checking, factory layout, logistics, procedural tasks, product and process design, and training. For each domain they selected one use case by consensus, judged against industrial relevance, technological feasibility and the added value of combining AR and VR, and produced wireframes and mockups of the envisioned interfaces.

The proposed concepts illustrate how CR could exploit the complementary strengths of the two technologies. AR excels at overlaying instructions onto real components, supporting pick-by-vision logistics, procedural guidance and early error detection, but it suffers from tracking inaccuracies on small or reflective objects, occlusion problems and reduced situational awareness. VR offers fully controlled, distraction-free environments ideal for simulation and hazardous-scenario training, but building complete virtual models of large facilities is labor-intensive and often blocked by missing CAD data. In the proposed logistics use case, for example, operators would preview a warehouse and plan a route in immersive VR, then execute picking with AR assistance, deliberately avoiding headset use while driving vehicles for safety reasons. In discrepancy checking, AR would compare a physical piping system against its digital twin, while VR would let users isolate components and validate checklists when real-world clutter overwhelms the overlay.

The experts also catalogued the obstacles standing between these concepts and deployment. CR research is still nascent, with no standardized design guidelines and few studies beyond procedural workflows. Transitional interfaces that let users move between AR and VR without disorientation remain an open research problem. Real-time visualization of complex industrial assets, synchronization of digital and physical environments, and the manual, time-consuming generation of virtual replicas are significant technical bottlenecks, though automatic model generation from CAD data and neural 3D reconstruction from multi-view captures could ease them. Hardware adds its own constraints: headsets are heavy, batteries limit session length, and many devices lack the intrinsic safety certifications required in safety-critical industrial settings.

The authors are careful to position their use cases as hypotheses rather than proven solutions. No functional system was built, and the design session insights are explicitly preliminary, intended to give the research community concrete directions for empirical validation. Future work, they write, should focus on developing functional prototypes with particular attention to transitional interfaces and novel interaction techniques, followed by comparative user studies in real industrial contexts, with continued collaboration with industry stakeholders to keep the concepts aligned with practical needs. They also suggest comparing optical see-through devices using dimming mode against video see-through headsets to determine which hardware best supports CR transitions.

So is cross reality crossing the industry? The honest answer from this study is that it is knocking but has not yet entered. The technology stack, from passthrough headsets to digital twins, is converging in ways that make seamless AR-to-VR transitions technically plausible, and early pilot results hint at real benefits for workflow efficiency, collaboration, workload reduction and usability. Yet with only 16 relevant studies in the scholarly record, no documented industrial deployments in the grey literature, and a predominantly exploratory evidence base, CR remains a promising research frontier rather than a mature industrial tool. For factories weighing investments in immersive technology, the message is to watch this space closely, because the transition between realities may soon become as routine as the transition between tasks.

Subject of Research: Industrial applications of Cross Reality combining augmented and virtual reality

Article Title: Is cross reality crossing the industry?

Article References: Marangelli, L., Laviola, E., Uva, A. E., & Gattullo, M. (2026). Is cross reality crossing the industry?. Journal of Ambient Intelligence and Humanized Computing. https://doi.org/10.1007/s12652-026-05130-2

Image Credits: AI Generated

DOI: 10.1007/s12652-026-05130-2

Keywords: Cross Reality, Augmented Reality, Virtual Reality, Industry 4.0, Head-Mounted Displays, Systematic Literature Review, Digital Twin, Factory Layout, Industrial Training, Human-Computer Interaction, Collaborative Workspaces, Immersive Technology