1a. Industry and logistics

Scope and introduction

Industrial Virtual Worlds (VW), using a variety of immersive technologies, have significant potential to transform industry and logistics. These digital environments, accessible on multiple devices, allow for the creation and use of Digital Twins. By leveraging these VW, businesses can simulate and analyse complex manufacturing and supply chain processes, improving forecasting, efficiency, and reducing costs. These technologies also support predictive maintenance, optimized workflows, and data-driven decision-making. As they evolve, their cross-platform capabilities foster collaboration and innovation, making them essential for the future industrial landscape in various domains.

Industrial processes comprise several areas where VW can confer significant benefits. These include design and engineering, virtual testing, planning and manufacturing, operations (including training), and maintenance, repair, and overhaul (MRO). Logistics is an intrinsic component of all these stages and processes. All these processes involve interactions both within a single company and across inter company networks. Consequently, ensuring that future industrial VW are structured on interoperable rather than proprietary platforms is a critical step to facilitate information flow across organisational boundaries.

The industrial VW market is experiencing considerable growth, with recent reports projecting a compound annual growth rate of approximately 35%, a notable figure compared to traditional technology sectors. Forecasts for the global market value in 2030 range from 100 billion USD (source: ABI Research³) to 395 billion USD by 2034 (source Precedence Research⁴) forthcoming decade. This trend is corroborated by studies indicating increased investments in VW technologies, with 62% of participants augmenting their spending from 2023 to 2024, and the number of large companies investing over 10 million USD in these technologies doubling within the same period. This trajectory underscores the rapidly increasing relevance of VW in industrial contexts. The adoption of VW in the industrial sector has notably accelerated, driven by concrete results and clearly defined use cases. Early deployments have demonstrated substantial improvements, including productivity increases of 15–20% through digital twins like in the Siemens Electronic Factory in Erlangen, reductions in product development cycles by 20–50%, and decreased downtimes. Additionally, virtual technologies significantly contribute to sustainability by reducing emissions through virtual prototyping, remote operations, and minimized physical resource usage.

The fundamental technology underpinning the effective utilisation of industrial VW is the integration of standardised DT/shadows/models. This applies to a diverse array of assets, including machinery, buildings, environments, processes, and materials. It is imperative not only to standardise the representation of real-world assets within virtual environments but also to standardise their interactions and connections with each other and with the real world. While various standards and normative efforts are currently underway, their development must accelerate and encompass a broader range of application fields to propel industrial VW forward. Different approaches can be employed, such as achieving interoperability through a combination of automated application programming interface integrations and a layered approach to standards, as detailed in the "Interoperability and Standardisation" chapter of the broader SRIA.

Use cases

This section outlines key research areas within the industrial and logistics domain, detailing their significance, current challenges, and proposed objectives for future innovation.

1a.1 Rapid prototyping through collaborative design, editing and simulations

This approach aims to enhance the cost and resource efficiency of prototype design cycles for designers and developers. By facilitating virtual experiences and testing of product and service versions and functions collaboratively with colleagues worldwide, it reduces the need for physical prototypes or expensive 3D printing. Furthermore, it enables secure co-development of innovative products and services with external partners and customers, thereby improving early-stage acceptance and feasibility. This method allows distributed development teams to collaborate on virtual prototypes, adapt designs in real-time using immersive technologies like Augmented Reality (AR) or Virtual Reality (VR), and interact with complex products or services at a 1:1 scale, with immediate visualisation of changes. This interdisciplinary approach can integrate various data and information layers of virtual models, providing shared and contextual experiences in realistic VW environments throughout the product lifecycle. It also supports interactive simulation and testing of all possible configurations with diverse customer groups.

Challenges and opportunities: Current practices often involve various development departments, frequently geographically dispersed, collaborating either virtually or on-site, but high transaction costs often lead to work division and limited collaboration. Prospective customers and end-users are typically engaged late in the development process, which can result in significant error costs and a risk of inadequate market acceptance. Currently, the adoption rate of VW solutions for rapid prototyping is low, with only initial proofs of concept and isolated lighthouse solutions available, indicating a limited presence of industrialised solutions in the market.

Research and Innovation Objectives: Industry-grade head-mounted displays and multi-user platforms (2a). Back-end platforms, integration of context and usage data, and utilisation of physical geometric virtual models (2d, 2e). Distributed-ledger technologies and AI for enhanced collaborative design and prototyping (2d, 2f).

1a.2 Virtual integration of factory planning and simulation of series start-up

This scenario aims to minimise the time required for production line start-up and reduce downtime for existing lines by enabling factory planners and systems engineers to plan and simulate production line modifications in advance. It offers a cost-efficient solution by saving time during the ramp-up and testing phases due to comprehensive prior simulation. Moreover, it mitigates the risk of errors and the need for re-planning, as diverse scenarios can be tested that are difficult to visualise and comprehend in a two dimensional environment. The outcomes can be made accessible on mobile devices, further enhancing flexibility.

Challenges and opportunities: While some very large companies already utilise VW for planning new facilities, access to the necessary know-how and, crucially, a sufficient DT base of production equipment and building information modelling models are not readily available for small and medium sized enterprises. The current adoption rate of VW solutions for virtual factory and production line planning remains low, primarily confined to a few very large companies.

Research and Innovation Objectives: DTs for comprehensive factory and production equipment (2e). Interoperability with enterprise resource planning systems (2d). End-to-end integration and interoperability among simulation tools (2d, 2e).

1a.3 Model-Based Systems Engineering (MBSE)

This approach aims to efficiently coordinate development teams, end-users, and other key stakeholders, thereby accelerating feedback loops and ensuring that project outcomes align with expectations. Virtual world-based project models, leveraging extended reality (XR) technologies, offer an accessible, highly visual, and interactive unified language applicable across different engineering disciplines. This facilitates a shift from conventional linear, rigid, and document-centric models towards a more collaborative, non-linear, and agile digital information-based methodology. Through fully immersive VR environments or context-aware AR overlays, teams can engage with project data in intuitive and flexible ways, significantly enhancing communication.

Challenges and opportunities: Engineering projects currently often revolve around various models that capture key aspects by highlighting important subsystems and elements while simplifying or omitting less relevant features. MBSE approaches exhibit considerable variation across different engineering disciplines, and the optimal method for conveying a model remains a frequent subject of debate. Despite the relative popularity of MBSE approaches, the utilisation of VR as a means of facilitating them remains underutilised, indicating a low adoption rate for VW solutions in this area.

Research and Innovation Objectives: Real-time 3D rendering engines (2c). XR head-mounted displays (2a). Collaborative multi-user VW for project coordination (2b).

1a.4 System control interfaces

This research topic focuses on enabling workers to remotely operate various machinery systems XR head-mounted displays facilitate a shared, real-time spatial context between remote workers and on site machinery, which is often crucial for effective remote control. This capability addresses challenges in many industrial contexts, such as production, medical surgeries, and construction, where integrating spatial information in real-time is vital for shared understanding among users. Remote operation reduces project costs and delays that can arise from the unavailability or high demand for key on-site personnel.

Challenges and opportunities: Current industrial practices frequently necessitate on-site presence and co-located interaction for system operation and control, which can lead to increased project costs and delays. While some -mainly- large companies are developing and employing XR solutions across their value chains, the adoption rate of VW solutions for system control interfaces is currently medium.

Research and Innovation Objectives: Advanced XR head-mounted displays (2a). Low-latency network connectivity to support real-time remote operation (2d).

1a.5 Manage and maintain intralogistics (AGVs, humanoids, forklifts)

This use case aims to minimise downtime and errors in intralogistics operations, ensuring that available resources are used efficiently. In photorealistic virtual environments, it is possible to train and simulate the behaviour of automated guided vehicles and humanoids. The interaction between tracked manual entities and robots like Automated Guided Vehicles (AGV) can be safely monitored, allowing robots to plan and simulate their future actions within the virtual world. Workers can visualise the real-time planned actions, safety zones, or trajectories of these robots using AR devices. In the event of errors, a rollback function with real track data can be used to locate missing goods within an interconnected system.

Challenges and opportunities: Currently, intralogistics operations are handled by a multitude of proprietary tools, each managing specific aspects such as automated guided vehicle guidance or warehouse maintenance. Manual labour, such as forklift transport, is often not tracked at all, and humanoids are largely confined to test cases rather than widespread production use. While some aspects like automated guided vehicle management or smart warehouses are well-established in two dimensional environments, interoperable and integrated VW solutions for the entire intralogistics fleet, encompassing training, planning, and predictive maintenance, are not yet available. The adoption rate of VW solutions in this area is medium, indicating a gap in comprehensive integrated solutions.

Research and Innovation Objectives: DTs for intralogistics assets (2e). Interoperability with enterprise resource planning systems (2d). 3D photorealistic training environments (2b). Advanced tracking and mapping technologies (2a, 2b). It is acknowledged that this use case benefits from collaboration with the ADRA partnership and that any double work is to be prevented.

1a.6 Training and onboarding in a virtual work environment

This research aims to enable employers to comprehensively train new employees and specialists at the earliest possible stage when new products and services are introduced, preparing them for new tasks. This ensures that theoretical concepts can be effectively implemented in practice. It particularly focuses on simulating critical operational situations in advance to enhance employee confidence and mitigate errors. VW provide flexible and experience-based learning opportunities for new activities and tasks. By simulating various scenarios, different courses of action can be tested and their consequences assessed. The inclusion of avatars and AI agents facilitates the training of social interactions and soft skills. Critical operations can be jointly trained by multiple participants in advance and on a case-by case basis, especially for situations in hazardous environments.

Challenges and opportunities: Currently, when new products and services are introduced, employees often receive training late in the process, on the job, or are provided with product sheets, process descriptions, and design plans that are difficult to comprehend. Critical situations or processes in hazardous environments are typically only discussed theoretically and cannot be experienced or simulated in advance. The adoption rate of VW solutions for training and onboarding is currently medium, indicating that while offers exist in the market, they are not yet broadly adopted across industries.

Research and Innovation Objectives: Advanced XR head-mounted displays (2a). Creation of DTs and human DT’s (2e). Collaborative platforms with the capability to manipulate and interact with virtual environments (2b, 2c). It is acknowledged that this use case benefits from collaboration with the ADRA partnership and that any double work is to be prevented.

1a.7 Virtual dashboards

This research focuses on enabling shopfloor managers to access relevant data and collaboratively monitor and analyse the shopfloor in a virtual world, utilising services and applications including AI. This approach ensures that data and DT of assets and processes are available to all authorised individuals, everywhere, with correct access rights. Beyond shopfloor managers, other stakeholders can also benefit from this data. Managers can inspect error cases within a virtual environment to gain a better understanding of bottlenecks, crashes, and other issues. This allows for real-time analysis from anywhere in the world, reducing downtime and facilitating effective communication of necessary changes to higher management by providing them with virtual experiences of the issues.

Challenges and opportunities: Currently, shopfloor managers access different enterprise resource planning or evaluation tools, typically displayed in tables, spreadsheets, or two-dimensional environments. The primary challenge lies in providing available shopfloor data in real-time within a virtual environment to anyone with access rights on demand. There are currently no widely adopted virtual applications for shopfloor integration that facilitate real-time, AI-assisted analytics on shopfloor data and collaborative access. The adoption rate of VW solutions for this purpose is low.

Research and Innovation Objectives: DTs and robust interoperability with enterprise resource planning systems (2d, 2e). End-to-end integration and interoperability of simulation tools (2d, 2e). Enabling real-time data access and real-time streaming capabilities (2b, 2d, 2e).

1a.8 Public outreach for energy grid extensions and generation units planning

This research aims to allow citizens to visualise planned power grid lines, new wind or photovoltaic parks, and power plants before construction, enabling them to understand the environmental impact on their living surroundings. For planners, it provides a means to effectively demonstrate planned assets to citizens, better address objections and concerns, and support and accelerate communication and public outreach during the planning process. This approach facilitates ad-hoc visualisation of planned assets in real environments on mobile devices, for instance, through applications. This offers citizens low-threshold access to dynamic visualisations and improves communication between citizens and planners during project planning and projection.

Challenges and opportunities: Current methods rely on static photoshopped images from specific viewpoints or animations in virtual environments, which offer a more abstract and less tangible impression of the actual impacts of new assets. These methods also lack the ability to provide visualisations from various viewpoints and under different conditions, such as weather or seasonal changes. The adoption rate of VW solutions for public outreach in energy planning is low, with only initial proofs of concept and isolated lighthouse solutions available.

Research and Innovation Objectives: AI for enhanced visualisation capabilities (2f). Mobile applications that integrate georeferenced planning data and virtual models of assets to be built.

1a.9 Life cycle continuity in structural and civil engineering projects

This research aims to provide infrastructure project stakeholders, such as architects, engineers, and construction managers, with a more comprehensive understanding of their project space and its main components throughout its entire lifecycle, from design and construction to maintenance. Multi-user, multi-location XR tools can be utilised to display relevant infrastructure models, enhancing spatial perception, improving communication, and reducing errors. Additionally, Mixed Reality (MR) and AR tools can be employed during actual construction and maintenance work to display relevant parts of the Building Information Modelling (BIM) model to field technicians and facility managers. The integration of a virtual assistant, such as a dedicated large language model, can provide real-time guidance, particularly during construction and maintenance phases, especially in non-nominal or unexpected situations.

Challenges and opportunities: Currently, a 3D digital representation of a project is developed and distributed among key stakeholders for tasks such as asset tracking, space planning, and Internet of Things data. While BIM is already well-established in the construction and infrastructure sector, the adoption of VW solutions built on top of this for lifecycle management is high for distinguished use cases. However, the full potential for immersive, collaborative experiences and advanced real-time guidance throughout the entire project lifecycle remains a research gap.

Research and Innovation Objectives: AR/VR/MR head-mounted displays and large XR displays (2a). AI and large language models for real-time guidance and assistance (2f).

1a.10 Virtual showrooms and product customisation in real-time

This research aims to enable sales representatives to clearly demonstrate innovative products and services and their benefits to potential customers. The objective is to help customers recognise the advantages and high quality of technically complex products and intangible services before making a purchase decision, thereby differentiating from competitors. VW allow for the presentation of virtual models of products and services within their virtual application environment, with concrete added value demonstrated using real data. In collaboration with customers, various application scenarios can be simulated, and, at least in part, immediate adjustments can be made.

Challenges and opportunities: Current methods involve elaborate presentations, trade fair appearances, films, and prototypes that often only abstractly illustrate the tangible benefits in a working or living environment. Communicating the direct benefits to a customer is particularly challenging for technically complex products, market innovations, and intangible services. Direct feedback and adjustments can typically only be made retrospectively. The adoption rate of VW solutions for enhanced customer experience is currently medium; while static virtual models for technical products exist, dynamic models or the ability to simulate usage in a customer's environment are limited.

Research and Innovation Objectives: 3D DTs with predictive and prescriptive analytics capabilities (2e). Physical places for presentation and interoperable virtual platforms.

1a.11 AI generated industrial Digital Twins and visualisations

This research aims to enable software engineers and 3D artists to generate 3D assets and DT of industrial machines and goods with the assistance of AI and scanning/mapping technology, thereby reducing manual development time and democratize the creation of DT. In an industrial virtual world, new objects could be placed on the fly with minimal integration effort. For example, an AI assistant could generate a 3D model based on available Computer-Aided Design (CAD) data from an enterprise resource planning system, cloud network, or marketplace, and then individually generate the DT of the necessary component, machine or process according to specific requirements. This process can be coupled with low-resolution scanners from widely available hardware, such as smartphones, to create digital assets from real environments with minimal effort and to make it affordable even for Small-Medium Enterprises (SMEs).

Challenges and opportunities: Currently, industrial 3D assets are typically created using CAD data and discipline-oriented software that can be operated only by few experts. Thus, the adoption is very limited and expensive. The challenge lies in integrating AI and large language model tools to generate assets with less manual integration work or even in automated workflow. Additionally, the current state of scanning and mapping technology needs improvement to build more detailed models more cost-efficiently. While tools exist that combine visual 3D models with software functionality, manual labour is still extensive. The AI generation of standardised DTs is still in an early research state, and while scanning and mapping technology is used for digitalising environments, it remains expensive for achieving detailed results. The adoption rate of VW solutions in this area is currently medium.

Research and Innovation Objectives: Integration of computer-aided design, AI, large language models, and 3D technology/game engines (2b, 2c, 2e, 2f). Development of advanced DTs. (2e). Advancement of scanning and mapping technologies (2d, 2e).

1a.12 Collaborative remote maintenance and operations support

This research focuses on allowing technicians and remote engineering teams to visualise machines and relevant equipment in high-fidelity 3D to support activities such as remote inspection and collaboration. This addresses the challenge that maintenance and operations support typically require on-site presence, which can lead to downtime and project delays. VR enables remote teams to visualise machines, equipment, or infrastructure in 3D, at full scale, and with high fidelity. This aids in maintenance planning, troubleshooting, and reviewing inspection reports that are difficult to interpret in two dimensions. Through AR or MR, on-site workers can receive real-time information from expert teams, even overlayed on real machines or components. Additionally, connection to a virtual assistant, such as a dedicated Large Language Model (LLM), can guide both the remote expert and the on-site worker in non-nominal situations. Human factors are crucial, ensuring that information is provided in the most suitable format, particularly for on-site workers.

Challenges and opportunities: Normally, maintenance and operations support necessitate on-site presence, which is not always feasible, resulting in downtime and project delays. While organisations in the industry are integrating VW into their workflows, for instance, to simulate and monitor complex factory systems, the practice still has considerable untapped potential. The adoption rate of VW solutions for collaborative remote maintenance and operations support is currently medium.

Research and Innovation Objectives: XR head-mounted displays and low-latency network Connectivity (2a, 2d). Consideration of human factors and User Experience (UX) in system design (2b, 3a, 3d). AI for enhanced support (2f).

1a.13 Exchange of complex industrial data

This research aims to enable manufacturing systems engineers to securely share DTs and Internet of Things data with suppliers and partners. This collaborative approach facilitates real-time optimisation of production, predictive maintenance, and supply chain operations. It allows for immersive and collaborative interaction with DTs within a shared virtual space, utilising blockchain-based access control for secure data sharing and decentralised digital asset ownership. Furthermore, it supports real time simulation and data fusion from multiple sources.

Challenges and opportunities: Currently, data exchange predominantly occurs through centralised platforms or manual file transfers (e.g., email, FTP). Proprietary platforms often exhibit poor interoperability, and there are significant concerns regarding security and intellectual property (IP) protection. Data silos hinder real-time collaboration, and there is limited traceability and access control. The adoption of VW solutions in industrial applications remains in an early stage, with implementation primarily observed in pilot projects and testbeds rather than full-scale rollouts. While DTs are more mature, collaborative environments are still fragmented.

Research and Innovation Objectives: Interoperable DT standards, such as Asset Administration Shell and DT Definition Language (2d, 2e). Decentralised data exchange mechanisms, including IPFS, blockchain, and Solid Pods (2d). AI and machine learning for semantic data translation (2d, 2f). XR interfaces for immersive collaboration (2b). Integration of edge AI and federated learning for privacy preserving computation (2d, 3b).

1a.14 Training AI systems and autonomous agents

This research focuses on enabling engineers and developers to train and test AI and machine learning models and autonomous systems, including robots and humanoids, within high-fidelity, multi-user virtual environments. These environments simulate complex physical and socio-technical systems, allowing for validation of performance, safety, and adaptability under diverse conditions prior to real world deployment. The proposed scenario leverages the availability of a DT of a complex system within a virtual world. This DT embeds various layers, encompassing realistic 3D visualisation and models that define its behaviour from multiple perspectives (e.g., physical level for process configuration and control, discrete event level for high-level behaviour and integration within a manufacturing system). This DT can generate synthetic data at scale under diverse and customisable conditions, facilitating robust AI training and testing. Robots and vehicles can be virtually deployed in detailed factory scenarios or urban environments, leading to reduced costs, safety risks, and faster time-to-market. Multi-user capabilities support interdisciplinary collaboration and scenario co-design.

Challenges and opportunities: Current training methods for AI and autonomous systems heavily rely on physical prototypes, limited-scale simulations, or domain-specific datasets. These approaches often lack scalability and struggle to fully account for diversity and safety issues, particularly in scenarios involving rare edge cases or high-risk situations. Existing simulation tools are frequently siloed, with limited multi-user interaction or cross-disciplinary integration. The adoption of VW solutions for this purpose remains in an early stage, with uptake primarily in academia or pilot industrial applications for very large companies. Barriers include high development costs, a lack of standardisation, technical complexity, and the advanced use of heterogeneous software libraries and tools.

Research and Innovation Objectives: High-fidelity 3D simulation engines (2c). DTs with real-time physics modelling (2e). Synthetic data generation pipelines and integration with real-world sensor and control data (2f). UX design for immersive collaboration (2b). Sim-to-real transfer and domain adaptation (2d, 2e). Development of governance frameworks for virtual experimentation (3b).

Recommendations

To advance the integration and impact of VW within the industry and logistics sectors, a series of strategic recommendations are proposed, addressing the identified challenges and research gaps.

Firstly, a concerted effort is needed to develop and promote interoperable standards for DTs and VW platforms. The current fragmentation, characterised by proprietary tools and siloed data, significantly hinders widespread adoption and seamless collaboration across supply chains and industrial networks. Standardisation initiatives, such as Asset Administration Shell must be accelerated and broadened to cover a wider range of industrial applications. This will facilitate secure, decentralised data exchange, crucial for real-time optimisation and predictive maintenance.

Secondly, investment in advanced hardware and software solutions specifically tailored for industrial applications is paramount. This includes the development of robust, cost-effective, and user-friendly XR head-mounted displays and large XR displays, alongside low-latency network connectivity. Furthermore, the creation of high-fidelity 3D simulation engines with real-time physics modelling is essential for accurate virtual testing, factory planning, and the training of AI systems and autonomous agents.

Thirdly, there is a critical need to enhance the accessibility and usability of VW solutions for small and medium-sized enterprises. This involves developing low-code or no-code content authoring tools that enable subject-matter experts to create and customise immersive environments without extensive programming knowledge. Furthermore, fostering knowledge transfer and providing clear demonstrations of measurable benefits will encourage broader adoption beyond large corporations.

Fourthly, the integration of AI and LLMs across the industrial VW ecosystem should be prioritised. This includes AI-assisted generation of DTs and machine visualisations, intelligent assistants for remote maintenance and operations support, and AI-driven analytics for virtual dashboards. These advancements will reduce manual development time, enhance decision-making, and improve the efficiency of complex industrial processes.

Finally, a strong emphasis should be placed on human factors and UX design. Ensuring that VW interfaces are intuitive, inclusive, minimise cognitive load, and provide information in the most suitable format is crucial for effective training, remote operations, and collaborative design. Research into user acceptance, long-term retention, and the transfer of skills from virtual to real-world contexts will be vital for successful implementation.