MAGIC: Multi-User Advanced Graphic Immersive Configurator for Sustainable Customization of Complex Design Products—A Sailing Yacht Case Study

Abstract

Modern design products are increasingly complex and emotionally significant, demanding versatile and collaborative customization. However, a literature and commercial review reveals a limited availability of flexible, multi-user, photorealistic Virtual Reality (VR) systems for product configuration. We introduce MAGIC (Multi-user Advanced Graphic Immersive Configurator), a collaborative platform combining realistic graphics with ergonomic validation using digital avatars, addressing the limitations of 2D visualization and existing tools. MAGIC is evaluated in a yacht design case study involving 30 participants in an immersive, co-located configuration of a sailing yacht to assess the system’s usability and the potential of VR for customizing complex products. Results show MAGIC’s feasibility in supporting multi-user configuration (100% success rate) and achieving a strong usability score (SUS = 80.83). User feedback highlights that high-quality graphics and additional content significantly enhance immersion and user engagement. However, encountered challenges with navigation methods and spatial perception indicate areas for improvement. MAGIC’s collaborative and immersive capabilities can be extended to other industries demanding proactive customer engagement in the customization of large, heavy products and ergonomic design. Moreover, by promoting prototype dematerialization and providing an interactive remote tool for end users, MAGIC offers potential environmental and economic benefits to boost the competitiveness of small and medium enterprises (SMEs).

1. Introduction

The product design industry is a high-value sector defined by style, quality, and customization. The evolving market landscape and advancements in industrial processes [1] demand solutions that actively involve customers in the configuration of complex products with emotional value.

VR configurators address these needs by providing a virtual platform to enhance product awareness by means of a flexible purchasing experience. These tools enable users to visualize, manipulate, and interact with complex designs firsthand [2], while also experiencing product functionalities interactively.

The spatial perception enabled by VR [3] has the potential to revolutionize how customers evaluate products, supporting industries that aim to convey comfort and ergonomic qualities, including automotive, interior design, camping equipment, and nautical applications.

VR also supports collaboration through shared and remote presence, enabling professionals, sales assistants, or other users (e.g., relatives or partners) to participate in the configuration process (Figure 1). This creates a more dynamic customization experience while providing foreign customers with remote tools to preview products realistically, eliminating the cost and inconvenience of international travel.

Additionally, the growing focus on eco-sustainability supports the adoption of low-impact remote solutions [4], enhancing the competitiveness of large, heavy products such as campers, boats, and houses produced by SMEs with limited logistics infrastructure and far from major trade fairs.

Despite VR’s potential, research on collaborative customization for emotionally impactful [5] products (e.g., yachts) remains limited; while a few VR solutions offer multi-user configuration tools with photorealistic visuals, these are often expensive and poorly optimized for mid-tier devices, underscoring the need for adaptive and flexible solutions to enable realistic previews through high-quality rendering.

To address these gaps, we introduce MAGIC (Multi-user Advanced Graphic Immersive Configurator), a collaborative cross-platform (VR/desktop) solution specifically designed for configuring complex, high-value products. MAGIC combines multi-user functionalities with optimized photorealistic graphics to deliver a scalable and immersive customization experience, even on lower-performance hardware.

MAGIC also advances research by providing a user study conducted in a real-world industrial context at a local shipyard, addressing the limitations of traditional 2D visualization systems that often fail to engage customers effectively. Through its immersive and intuitive interface, the system bridges the gap between technical CAD designs and customer expectations, offering a practical, accessible, and efficient solution for industries where emotional engagement and collaborative design are essential.

Overall, this research contributes to the scientific literature and promotes eco-sustainability by achieving the following:

• Developing MAGIC, a novel VR collaborative configuration system with real-time optimized photorealistic rendering;
• Providing a case study of MAGIC in a shipyard as a real-use scenario and gathering impressions from end users.

2. Related Works

State-of-the-art VR multi-user configuration is addressed on two levels, one in the scientific literature and the second in a commercial context.

2.1. Scientific Literature

Scientific research addresses multi-user collaboration predominantly from a social perspective, emphasizing the role of VR in design review and group decision-making dynamics, ultimately leading to greater efficiency in the process. For instance, in the textile sector, VR/AR platforms have proven to significantly accelerate design workflows [6], while Mixed Reality (MR) facilitates collaborative prototyping among designers, ergonomists, and end users [7].

Recently, ref. [8] introduced CIDER, an AR system that enables independent and conflict-free editing of shared scenes, analyzing the impact of “commit” strategies (forced vs. consensus-based) on collaborative dynamics and usability. In addition, ref. [9] examined how user roles and task complexity influence subjective perception while also exploring technical factors such as latency, which affects Quality of Experience (QoE) in long-range virtual collaboration.

In the nautical industry, a noteworthy example is BoatAR [10], an AR-based configurator that simplifies remote boat customization. BoatAR has proven to reduce logistics and inventory costs while showcasing the potential of immersive technologies to address operational and economic challenges. Moreover, ref. [11] addresses the functional arrangement of cockpit instrumentation in maneuvering areas, supported by evaluations of operational, ergonomic, and anthropometric aspects.

Despite these advancements, multi-user product configuration in VR remains underexplored, particularly regarding eco-sustainability and the role of advanced graphics in enhancing emotional engagement. Persistent challenges include high hardware costs and the lack of standardized user interfaces to ensure accessibility and intuitiveness for a broad audience [12]. Moreover, despite technological progress, the inherent complexity of VR tools often requires specialized expertise for effective integration into complex workflows, such as those typical of the nautical industry [13].

Future research should explore novel collaborative approaches that incorporate realistic graphics and avatar support, enhancing spatial perception and ergonomic assessment in real-world industrial scenarios.

2.2. Commercial Solutions

We conducted a Google search with Gemini artificial intelligence (AI) [14] to identify the most significant commercial products. The query was “a list of software responding to the keywords Configuration, VR, Product Review, and Multi-user”. The results were then reviewed and filtered to narrow the search to the most relevant solutions.

Five software programs were found: Autodesk VRED [15], EyeCad VR [16], Keyshot Studio VR [17], Mindesk [18], and VR Sketch [19].

A technical analysis was conducted to assess the operational focus of these applications, evaluating the following parameters:

• Accessibility: Navigation method, quick-access PoV and UI user-friendliness
• Configuration: Management of aesthetic and geometric variants.
• Object Manipulation: Real-time tools for geometry modification and customization.
• Interaction and Inspection: Integrated tools for purposes such as measurements, clipping planes, annotation, and screenshots.
• Immersiveness: Visualization method, animations, and dynamic environmental features.
• Multi-User Capability: Technical requirements for collaborative configuration and communication.
• Synchronization: VR-CAD bidirectionality.

The following is a summary of the commercial analysis:

VRED Pro (Autodesk) [15]: A prototyping tool for realistic, collaborative scenes with advanced variant management and inspection tools (e.g., clipping planes, measurements). While effective, it lacks intuitive workflows for configuration and VR object manipulation, and it requires costly licenses and high-performance hardware for optimal multi-user sessions.

Mindesk [18]: A VR plug-in for CAD software enabling real-time object editing with bidirectional synchronization. Intuitive for navigation but lacking collision detection and configuration tools, it prioritizes technical precision over immersive rendering and offers limited support for product customization.

EyeCad VR [16]: A standalone software for immersive design approvals, providing customization features, intuitive navigation, and realistic rendering with a wide material library; while strong in interactivity, it lacks collaborative features and inspection tools.

Keyshot Studio VR [17]: A Keyshot extension for VR visualization, offering advanced configuration tools, precise geometry manipulation, and photorealistic rendering. High-end hardware and expensive licenses limit its accessibility for collaborative sessions.

VR Sketch [19]: Integrated with SketchUp, it enables synchronized CAD modeling of complex geometries along with dynamic tools such as section planes, measurement, and precise scaling. Although not optimized for photorealistic rendering, it supports mid-tier hardware and extends CAD tools effectively into VR.

A comparative table of commercial software aligned with the scouting parameters mentioned above is presented here (

Table 1. Comparison of VR software features.

	VRED Pro	Mindesk	EyeCad VR	Keyshot VR Studio	VR Sketch
Accessibility
User Locomotion	Teleport	Fly Mode	Teleport/ Walkthrough	Teleport/Fly Mode/Walkthrough	Teleport/Fly Mode/Walkthrough
Viewpoints	✓	×	✓	✓	✓
UI	Intuitive	Limited	Limited	Intuitive	Versatile
Configuration
Aesthetic Variants	✓	×	✓	✓	✓
Geometric Variants	✓	×	×	✓	×
Object Manipulation
Move	×	✓	×	✓	✓
Rotate	×	✓	×	✓	✓
Scale	×	✓	×	✓	✓
Grab and Drop	×	✓	✓	✓	✓
Interaction and Inspection
Mechanism Simulation	✓	×	✓	✓	×
Measurement	✓	✓	×	✓	✓
Exploded	×	✓	×	×	×
Section Planes	✓	✓	×	✓	✓
Show/Hide	×	✓	×	×	✓
Annotation	✓	✓	✓	✓	✓
Screenshot	✓	✓	×	✓	✓
Immersiveness
Visualization Method	Ray Tracing/ Baked Lights	Shadowed	Baked Lights	Ray Tracing/ Baked Lights	Shadowed
Animations	✓	×	✓	✓	×
Lighting	✓	✓	✓	✓	✓
Scenario	✓	Limited	✓	✓	Limited
Additional Contents	✓	Limited	✓	✓	Limited
Multi-User Capability
Sharing Requirements	High	Moderate	×	High	Low
Live Chat	✓	✓	×	✓	✓
Synchronization
VR ↔ CAD	×	✓	×	×	✓

).

The tools described above showcase diverse approaches to VR-based design and customization, but they also reveal shared challenges. VRED and Keyshot Studio VR prioritize rendering quality and advanced configuration tools, yet their inability to optimize graphical resources limits their accessibility on mid-tier hardware. Additionally, their complex workflows and steep learning curves complicate virtual scene development, leading to dependence on proprietary authoring processes. Multi-user collaboration further amplifies these challenges, as all participants must own the same license package as the host to access full scene functionalities, increasing costs and reducing accessibility.

In contrast, Mindesk and VR Sketch focus on integrating VR into CAD workflows, enabling real-time co-design with synchronized files. However, their lack of dedicated tools for product customization and limited rendering quality make them less effective for detailed product evaluation. Moreover, VR Sketch collaboration is restricted to scene exploration, as editing functions are disabled. EyeCad, despite its user-friendly design and visual appeal, lacks synchronous collaborative features and inspection tools, limiting its versatility.

MAGIC seeks to address these limitations by combining key features into a comprehensive configurator. It ensures equal access to functionalities regardless of user authority within the scene; dynamically adapts rendering quality to match available hardware through a dedicated graphics management system; and delivers an engaging, emotional user experience. By overcoming hardware constraints, collaboration barriers, and customization challenges, MAGIC offers an inclusive and high-performing solution for VR-based design and configuration.

3. Implementation

We developed MAGIC to provide proprietary and specialized software for the collaborative configuration of design products. To implement collaboration, the users are embodied as digital avatars and provided with synchronized tools. A maximum of 10 participants can simultaneously use the following interactive features (Viewer Menu, see Figure 2 left):

• Transform: Allows users to select an object or assembly prepared for mobility, drag in 3D space, and reposition it.
• Measurement: Allows users to take measurements to support space evaluation.
• 3D Cut Volume: Allows users to model, move, rotate, and scale a cutting volume to dynamically inspect cross sections of the model.
• X-Ray: Allows users to select objects to make them transparent, enabling inspection of hidden geometries (Figure 2 right).
• Bookmark: Allows users to create quick-access points in the scene, facilitating access to key elements of the model.
• User Scale: Manages the avatar scale, allowing users to access an overall view of the model or move quickly through space.
• Annotation: Allows users to create 3D text boxes for comments and annotations.
• Snapshot: Allows users to take snapshots to validate the arranged configuration.

Unreal Engine 5 (UE5) was chosen for the VR implementation due to its versatility, scalability, and advanced integrated programming tools. With technologies such as Nanite and Lumen, this game engine can provide high-quality visual experiences and manage complex, high-resolution geometries with dynamic lighting.

Graphical settings such as Single-Pass Stereoscopic Rendering and Stereo Foveation are enabled in the project to optimize the VR experience, improving system fluidity by halving the computational load on each lens of the HMD and by focusing the maximum resolution at the center of the user’s field of view (FoV).

MAGIC is an extension of Collab Viewer, a cross-platform template provided by UE5 and designed for collaborative experiences, with excellent flexibility for integrating custom controls through Blueprint (BP) programming.

A dedicated BP logic has been implemented to allow the configuration of aesthetic variants, supported by photorealistic materials and algorithms for simulating interior furniture components, aimed at validating the ergonomics of the yacht’s confined loft spaces.

Using a ray-cast targeting logic and activating the command with the dedicated button on the controller, the user can easily cycle through material variants (Figure 3) and interact with the yacht’s internal and external components, without the need for an overlaid menu. This approach was chosen to prioritize the usability and simplicity of the interface.

In the scene, the user can move using the teleportation metaphor (Figure 4 left) or quick access points (Bookmarks) to explore the model and access an off-board PoV for an overview of the configured yacht. (Figure 4 right).

At the start of a collaborative session, the initial user (Host) enables the server, allowing participants (Clients) to access the shared scene. Each Client, equipped with an individual headset and computer, can join the session remotely by connecting to a shared network (Wi-Fi or VPN).

In basic multiplayer solutions, only the Host has the authority to edit the scene on the server and share it with Clients. Although default template features are provided with appropriate programming measures, implemented functionalities (e.g., material changes, animations, etc.) require tailored multi-user logic to allow all users to actively interact with the scene during the session.

To address this, we developed dedicated algorithms for Replication and Multicast to support multi-directional synchronization between the actions of the Host and the Clients. This ensures a seamless collaborative experience with reduced latency and no constraints related to user authority.

When a command is activated, the server evaluates in real time the authority of the user who triggered the function, following two distinct logical paths (Figure 5):

• If the user is the Host, the server automatically transmits the command to all participants, and the algorithm is executed in multicast.
• If the user is a Client, the remote VR Pawn is temporarily promoted to Host, receives authorization to access the server, and can then transmit the algorithm in Multicast to the participants.

Immersive elements such as spatial sound effects and dynamic environmental content (e.g., waves, boats in motion (Figure 6)) can be added to enhance emotional engagement and the overall quality of the experience. By means of volumetric sound occlusion, the pitch of waves and motor noises is reduced when an object is positioned between the sound source and the avatar, allowing users to experience a quieter and cozier environment once inside the loft. In addition, automatic exposure is enabled in the scene to ensure a smooth and realistic transition between the yacht’s interior and exterior, where sunlight might otherwise appear overly bright or dazzling.

Lastly, users can adjust the rendering quality by means of a graphic management tool designed and implemented by our team to access the Game User Settings (GUS) directly from the launch menu. This feature ensures that MAGIC is compatible with both current and future hardware with varying graphical capabilities.

4. Case Study: Racer/Cruiser Yacht Configurator

As a case study, we selected an Italian shipyard specializing in high-performance carbon fiber sailing yachts designed for recreational cruising but also capable of racing. This represents an ideal showcase for the potential of MAGIC, as these yachts are large, difficult to transport, and highly customizable in terms of aesthetics and technical elements and because they require careful configuration of limited spaces.

A focus group with the company’s head and the technical and commercial staff highlighted the need for virtualization tools to reduce transportation costs and environmental impact associated with trade fairs and client visits. A simulation was conducted to estimate the ${\mathrm{CO}}_2$ emissions related to transferring a ±13 m yacht from Bari to key fairs in Düsseldorf and Genoa (Figure 7).

We considered road distances of 1745 km and 960 km, respectively, and a total transport weight of 20.1 tons, including the yacht (5.6 tons) and the heavy-duty transport vehicle (14.5 tons). The estimated weight of the special transport vehicle is based on typical configurations for heavy-duty trucks and trailers used in oversized transport. This includes a tractor unit (8–12 tons) and a specialized trailer (2–8 tons; volvotrucks.com (accessed on 15 January 2025)). Assuming fuel consumption based on weight (The International Council on Clean Transportation (ICCT) reports that the fuel consumption of a typical tractor–trailer is approximately 0.326 L/km (theicct.org (accessed on 15 January 2025)). Based on a typical payload, fuel consumption per ton per km can be reasonably estimated at 0.4 L/ton/km, though this varies with vehicle efficiency, load, and road conditions. A widely accepted value for average ${\mathrm{CO}}_2$ emissions per liter of diesel is 2.68 kg ${\mathrm{CO}}_2$/L (drivingtests.co.nz (accessed on 15 January 2025)). On this basis, we estimated a fuel consumption of approximately 28,060 L for the Bari–Düsseldorf round trip, resulting in 75.2 tons of ${\mathrm{CO}}_2$ emissions. Using the same calculation, the Bari–Genoa route produces approximately 41.4 tons of ${\mathrm{CO}}_2$.

Results indicate that this transportation process generates approximately 116.6 tons of ${\mathrm{CO}}_2$ annually for one round trip to each destination. These findings highlight the importance of exploring more sustainable transportation methods or implementing compensatory measures to reduce the environmental impact of these operations.

Alongside this issue, customers—generally non-Italian—ask for a user-friendly remote configuration tool to avoid the time, cost, and burden of international travel. A configurator provides a more realistic idea of the result, enabling a more engaging and confident decision-making process.

Customer involvement currently relies on static 2D renderings and regular visits to monitor construction progress and support key decisions, often aided by disposable rapid prototypes, samples, and models. However, 2D desktop tools fail to effectively convey the perception of onboard spaces, lack ergonomic immersion, and prevent active interaction with environmental elements. For instance, the commercial team reported losing a large client because they could not fully demonstrate the accessibility of the final result.

An innovative full-carbon yacht design was chosen to showcase MAGIC’s capabilities. We imported the CAD models from Rhinoceros and modeled missing parts and components, materials, and setup. All variants available to define the aesthetic and functional aspects of the yacht’s features (e.g., hull, deck, rigging, furniture, lighting) were implemented to ensure flexibility in customization and immersive support to evaluate spaces within the cabin and on the deck.

In addition, we designed and implemented realistic and dynamic environments, such as breathtaking marine scenarios, which included open seas (Figure 8), ports, and coastlines, to enhance user engagement and create an emotional impact during the purchase process.

We address the following Research Questions (RQs) with a user test in a controlled setting:

• RQ1: Does MAGIC ensure an adequate level of usability?
• RQ2: Can MAGIC enhance sailing yacht configuration and provide realistic spatial perception?
• RQ3: Do advanced graphics impact immersion and realism for the configuration process?
• RQ4: Do environmental scenarios and additional content contribute to user engagement?
• RQ5: Do demographic data or prior VR experience affect the system’s usability?

4.1. Participants

Participants were recruited through a social call distributed via the university mailing list, including students, PhD candidates, and researchers from the Engineering and Industrial Design faculties at the Polytechnic University of Bari as a representative sample of users with different backgrounds for evaluating the system’s usability.

Thirty volunteers participated in the tests (age of 24.20 ± 4.62, min = 18, max = 32), with a gender distribution of 63.3% male (19 users) and 36.7% female (11 users). The male prevalence is in line with the current preference of real clients, while the age is younger to provide insights into future customers. To create a multi-user scenario, the subjects were divided into 15 pairs according to booking order, with the members of each pair not necessarily knowing each other.

4.2. Setup

The tests were conducted in a controlled environment at the VR3Lab research laboratory of the Polytechnic of Bari using Oculus Meta Quest 2 and 3 HMDs connected to two side-by-side laptops with different technical specifications:

• An Asus ROG with an Intel Core i7-7700HQ processor, 16 GB of RAM, and an NVIDIA GeForce GTX 1070 graphics card;
• An Acer Aspire 5 with an AMD Ryzen 7 5700U processor, 16 GB of RAM, and an AMD Radeon RX Vega 8 graphics card.

These two different setups enabled compatibility with devices having varying system specifications and operational conditions. Specifically, the Acer laptop is the same hardware used for scouting commercial software and can provide valuable insights into the performance offered by MAGIC compared to existing tools. Its specifications fall short of the recommended requirements for both professional software (e.g., VRED or Keyshot) and UE5.

4.3. Procedure

The test duration of 25 min was estimated, according to the company’s experience, as sufficient to perform an initial configuration of the yacht. Before the test, users were introduced to the research topic and provided with training (5 min) on the VR commands, with the total duration of the training and test being set to 30 min. Users, however, could interrupt the test at any time. The participants were asked to role-play as potential customers of the yacht model in an interactive and consensual configuration (Figure 9 left). During the collaborative session, users were co-present in the same laboratory and connected by two synchronized sets of hardware to a shared scene through a Wi-Fi network (Figure 9 right).

The users defined materials, nautical components, and interior furniture; analyzed spaces using the inspection tools; and interacted with dynamic elements of the scene to explore the model and assess the ergonomic suitability. During the test, a scientist monitored users’ actions to ensure safety and provide assistance.

4.4. Metrics

The configuration experience for each pair enabled the collection of data through the following metrics:

• User age, gender, and previous VR experience (yes/no);
• Success rate: Measures the effectiveness of the VR configuration process in terms of the percentage of successful paired sessions;
• Early exit: Number of pairs who quit the session before the planned test time;
• SUS [20]: Standard questionnaire for evaluating system usability;
• Open-ended feedback: Evaluation of the implemented features, the level of immersion, the graphical rendering, and the emotional impact, including any improvement suggestions.

The analysis of demographics such as age, VR experience, and gender may provide valuable insights into factors influencing usability. Below is a concise explanation of their potential impact:

• Age: Older users, who may resemble the actual clientele (e.g., boat owners), could prioritize configuration feasibility from a purchasing perspective, while younger, tech-savvy users might offer insights to improve features and enhance gamification.
• VR experience: Experienced users might focus on advanced features and provide technical feedback on immersion and interactivity, while inexperienced users could highlight accessibility gaps and the learning curve.
• Gender: Gender might influence interaction preferences, such as task approach and priorities, ensuring that the system design is inclusive and appealing to all user groups.

5. Results

Most of the participants (63.3%) reported having no prior experience with virtual reality, while the remaining participants indicated that they were familiar with VR (36.7%).

5.1. SUS

• Obtained SUS score: 80.83, above the acceptable threshold of 68.
• Distribution: 83.3% of participants assigned scores above 68, placing the application in the “acceptable” usability category (Figure 10).

5.2. User Feedback

User feedback (detailed in the appendix; see

Table A1. Participant feedback summary with positive comments and suggestions for improvement.

Participant	Positive Comments	Suggestions for Improvement
P1	Graphics and interaction with objects were good.	Preferred more materials for customization.
P2	Realism, navigation, and details were interesting.	Teleportation issues; suggested adding land/island context; suggested enhancing sail textures.
P3	The scene felt real and well designed.	Desired better interaction with the environment.
P4	Liked object interaction and independence.	-
P5	Intuitive, immersive experience.	Difficult to select smaller objects from a distance.
P6	Object representation enhanced configuration.	Suggested boat motion for added interest.
P7	Interaction and fluid animations were appreciated.	Reflections seemed unrealistic.
P8	Graphics suited product design.	Pointing system could be more precise.
P9	Immersive for VR beginners.	Teleportation sometimes failed.
P10	Graphics exceeded expectations.	Experienced fatigue and dizziness.
P11	Enjoyed interactive elements.	Moving often resulted in passing through boat walls.
P12	Positive interaction with elements.	Unsatisfactory color range; suggested an interface for selecting shades.
P13	Integrated details felt realistic.	Suggested adding haptic feedback, implementing a collision script, and addressing lag.
P14	System was engaging and responsive.	Left-handed controls were challenging for right-handed users.
P15	Appreciated customization and intuitiveness.	Teleportation often triggered the menu accidentally.
P16	Felt like objective reality and was easy to use.	Suggested a greater variety of colors.
P17	Highly realistic immersion; useful for decision-making.	Suggested a command guide for better function understanding.
P18	Simple main controls made the experience enjoyable.	Desired more freedom of movement inside the boat.
P19	Optimized interface; provided 360-degree view.	-
P20	Fantastic, lifelike experience.	-
P21	Enjoyed opening doors, picking objects, and moving freely.	-
P22	Excellent movement fluidity and realistic images.	Teleportation issues were noted.
P23	High-quality graphical rendering in an immersive environment.	Interaction was complex and needed simplification.
P24	Realistic setting with simple controls.	Menu occasionally closed unexpectedly.
P25	Liked real-time product change observation.	Not satisfied with limited fluidity of movement.
P26	Realistic textures and animations.	Teleportation sometimes worked poorly.

) provides a deeper understanding of the system’s strengths and weaknesses and highlights key areas for improvement. Four main themes are identified: Graphics and Immersion, Interaction and Customization, Navigation and Controls, and Functional Suggestions.

• Graphics and Immersion: Render quality and immersiveness were among the most appreciated aspects, with users emphasizing the high level of detail contributing to the experience. One participant remarked, “The graphics exceeded my expectations… it is an excellent way to customize and imagine your boat before purchasing it”. Many users felt they were truly aboard the yacht, thanks to the high-quality visuals of objects and materials. One participant noted, “The integrated details were so well done, it felt like I was actually inside a yacht…”. However, some users suggested improving material reflections and expanding the environment to make it more engaging (e.g., adding a nearby island).
• Interaction and Customization: The ability to interact with, customize, and move objects in real time was widely praised for being intuitive, making the experience practical and independent. Another participant stated, “I particularly liked interacting with the objects and making changes on my own, without needing someone else’s help.” However, users desired a wider variety of materials, colors, and textures with more realistic and practical options.
• Navigation and Controls: Although the controls were generally intuitive, seven out of thirty participants expressed a desire for greater freedom of movement and alternatives to teleportation, which was often inaccurate or challenging in tight spaces. As reported by one user, “… I was disappointed by the teleportation method which occasionally failed to recognize the correct floor to move on.” Additionally, selecting smaller objects remotely with ray casting proved to be challenging. Furthermore, some right-handed users found the placement of the configuration controls on the left controller uncomfortable.
• Functional Suggestions: Other suggestions for improvement included a more precise locomotion system, clearer training, and haptic feedback to indicate an avatar–object collision. Two out of thirty participants also reported physical fatigue and dizziness caused by the headset weight and occasional lag.

6. Discussion

With a success rate of 100%, every pair of participants agreed upon a final configuration within the set test duration. This outcome highlights the feasibility of the system in supporting collective decisions in a dynamic and immersive environment. Only one pair quit the session before the estimated time, as one of the participants needed to attend a university class. Therefore, it can be considered a statistical outlier, unrelated to any system interface design issue or episode of cybersickness, supporting the assertion that the 25 min session length was properly estimated.

To further investigate the distribution of individual SUS scores, potential correlations were analyzed to provide a comprehensive examination of their relationship with sample parameters such as age, gender, and prior experience with VR devices.

No significant correlation was found between users’ demographic data and the distribution of usability scores, which appeared to be largely random, thus ruling out a functional dependency between these two parameters (Figure 11).

The slope (r = −0.024) of the regression line indicates a weak negative correlation, showing that as participants’ age increased, the SUS score tended to decrease slightly.

However, since the p-value = 0.899 >> * p (<0.05), it can be concluded that this correlation did not hold statistical significance.

Similarly, the correlation between gender and SUS scores was negligible (Figure 10 left), with a percentage difference of 1.38% between the average values assigned by each category. Remarkably, the scores of male participants exhibited a greater deviation.

Regarding VR familiarity (Figure 12 right), experienced users provided scores within a broader range than non-experienced participants, even though the average results were similar (percentage difference < 0.3%). This might indicate that experienced users conducted more technical and informed evaluations, while inexperienced participants offered valuable insights into the system’s impact on end users without specific expertise.

The feedback indicates that the experience was generally perceived as immersive, engaging, and well executed, with a graphical and interactive system that exceeded users’ expectations. However, targeted improvements in customization, navigation, and technical optimization could address the issues that emerged, making the overall experience even smoother and more satisfying. With a greater variety of options, a more intuitive control system, and a more dynamic environmental context, this application could reach an even higher level.

Following the analysis of the results, it is possible to attempt to answer the RQs:

• RQ1: Supported—MAGIC’s usability falls within the “acceptable” range of the SUS metric.
• RQ2: Partially supported—MAGIC is feasible for sailing yacht configuration, but improvements in navigation and collision systems are required for effective space evaluation.
• RQ3: Supported—Photorealistic graphics are essential for a convincing digital preview.
• RQ4: Supported—Additional content significantly enhances the experience.
• RQ5: Not supported—Demographic data and VR experience do not significantly affect the SUS score, but further investigation with a wider sample is needed.

7. Limitations and Future Works

Tests revealed several limitations but also revealed opportunities for improvement that could guide more targeted and informed future development.

Regarding the user test, the university participants, while providing valid usability results, present a limitation in terms of sample diversity. In addition, the sample had a relatively young average age, highlighting the need for a broader demographic range along with better gender balance. Future studies could aim to include potential yacht customers in real-case purchase scenarios, and pair relationships could also be further explored as an alternative to the current semi-random pair generation approach.

As far as navigation, the current teleportation method enables rapid exploration by allowing users to reach any point in the scene and then move in real space within physical boundaries. However, as noted by some test participants, it often results in a fragmented experience and poses challenges for less experienced users due to its reliance on dexterity and familiarity with the software. Joystick locomotion can be a valid alternative to simulating walking. Nevertheless, this technique is prone to inducing motion sickness, as noted in the literature [21], and presents challenges when navigating tight spaces, such as the interior of a boat, due to collisions. Introducing a parametric avatar tailored to the user’s height and body type could support the implementation of alternative locomotion methods, offering a more natural and immersive experience.

In terms of performance, professional software has shown significant rendering limitations resulting in an impractical user experience. Under identical hardware conditions, MAGIC performed reasonably well despite minor fluidity constraints. Future tests will incorporate quantitative metrics (e.g., GPU consumption and FPS monitoring) to provide a comprehensive efficiency comparison. The co-located setup was requested by the partner company’s marketing staff. In their experience, a common client pattern involves couples either visiting the company or joining remotely, with no specific need for a far-apart configuration. Nevertheless, future work may also investigate latency and bandwidth parameters to assess the feasibility of long-range configurations.

Moving forward, a promising prospect is the integration of shared co-design tools within a virtual space, where clients and designers can collaborate in real time to create ad hoc solutions beyond pre-configured options. Users would participate in designing interior layouts and structural components, leading to a more proactive and engaging process. However, achieving this will require overcoming technical challenges, such as synchronizing the VR environment with CAD files and finely managing objects within the scene.

With these developments, MAGIC could evolve into an even more flexible and powerful tool capable of meeting the technical needs of designers and the expectations of end users.

8. Conclusions

This study examines the applicability of Industry 4.0 technologies in complex product configurations. Despite growing interest, only a few studies and commercial products have investigated multi-user tools for remote configuration purposes. Moreover, the potential of advanced graphics to enhance emotional engagement remains relatively unexplored, while existing software solutions often lack optimization and affordability.

To overcome these limitations, we developed MAGIC, a multi-user VR standalone prototype configurator, leveraging the advanced graphical and programming capabilities of UE5.

As a case study, MAGIC was applied to the sailing yacht sector to contribute to recreational boating by providing a VR solution aligned with the needs of the partner company, offering collaborative functionalities, immersive graphics, and virtual support for spatial evaluation. The project also integrated eco-sustainable and social aspects through the dematerialization of concepts and supported the competitiveness of the manufacturing sector.

Validation confirmed the software’s high usability, with a SUS score of 80.83, above the acceptable threshold. User feedback highlighted strengths and areas for improvement, providing valuable insights to further enhance the configuration experience.

The research demonstrated the feasibility of VR for realistic yacht configuration, although challenges with navigation methods and spatial perception remain to be further investigated. Despite this, VR technology is not yet mature enough to effectively integrate into industrial and design processes. However, as innovation progresses, these tools will become increasingly relevant, accessible, and high-performing.

As a final remark, MAGIC provides a foundation for future research projects in other industrial fields, with promising positive economic, commercial, and environmental impacts.