Simplify 3D modeling through AI
As we have discovered in recent months, generative AI has the potential to revolutionize the production of text content, images, and even music. The latest innovations leave no doubt that this technology is set to have a bright future in the world of 3D creation.
Each immersive learning production involves a preliminary step of 3D creation, which can vary in complexity. This may involve reproducing an entire industrial site, a warehouse, or a digital twin of a complex machine, for example. Our team is, of course, constantly monitoring solutions that will simplify or accelerate this step.
Audace has thus become acquainted with the company Mazing. Mazing develops a promising SaaS solution that uses AI to generate 3D models from photographs.
Exhibiting in the startup hall, this company is still in its early stages. At this point, it remains limited in the range of 3D models it can generate. Indeed, for an AI algorithm to be effective, it must be trained on a large volume of data. This is often a critical step, or even a technical bottleneck, for deploying such a solution. The needs related to the simulators developed by the agency are very specific (tools particular to a profession, rare or sometimes unique machines). Thus, they are not yet part of the recognized elements and may not be supported for some time. The first use case presented by Mazing is rather consumer-oriented: virtual fitting of clothing, including shoes.
Audace remains attentive to this sector, which is evolving at an impressive pace.
Use authoring solutions
The no-code trend (using tools without code) has significantly contributed to the emergence of authoring tools in recent years. These tools, for example, enable trainers to design their own educational content.
Those who have tested these solutions know that they offer undeniable gains in autonomy for the trainer. However, it is important to keep in mind that one cannot necessarily become an instructional designer or artistic director on a whim! The creations produced with these solutions can indeed result in very standardized or tedious content.
The Spectral TMS Case: An Example of an Authoring Solution for Augmented Reality Maintenance
CREATING OPERATING MODES AND EXECUTION ON THE FIELD WITH AUGMENTED REALITY
The goal of this solution is to assist in creating technical operating procedures using an authoring interface, without requiring specific skills in Unity or Unreal Engine. The content generated is then sent to the augmented reality headset of maintenance operators to assist them in real-time on the field in front of the machine to be repaired.
As part of the demonstration at the booth, the Audace team, equipped with Hololens 2 headsets, was able to carry out a task: diagnosing and repairing a fault observed on a PC tower. All of this was achieved without the need for computer expertise!
To do this, the operating procedure detailing the diagnostic steps and corrective actions to be implemented was prepared in advance using the authoring solution provided by the company.
Once on site, the headset first locates the actual PC tower using a QR code attached to it. The headset then displays guidance on the steps of the operating procedure to follow (such as checking certain parts, unscrewing indicated screws, etc.).
The information is not solely top-down. The maintenance technician using the tool always has the option to send field data back to the server. They can enrich the knowledge base with their feedback and observations (photos, comments, suggestions for improvements, etc.).
Our Experts Opinion: The solution appears very promising and user-friendly. It is clear that there are many potential use cases in the industry. It works well, in a simple and intuitive manner, for the end user on the Hololens 2. However, it remains limited to the use of this headset. A version available on smartphones is also offered but, unfortunately, provides less accurate 3D localization. New mixed reality headsets (such as the Meta Quest Pro) are not yet compatible with the solution, but their anticipated support could make the solution more accessible at a lower cost than the Hololens.
