Issues of Scaling 3D Content in Practice

Why emerging 3D cloud engines are important

8 minutes read

Examples of 3D Productions

Wayfair

Wayfair, a leading e-commerce platform for home goods, has long recognized the value of 3D visualization for delivering engaging shopping experiences. In their experience 3D content will serve as the foundation for visual merchandising as well as innovative applications such as AR and VR.

To prepare for the long run Wayfair in 2018, Wayfair launched a program designed to educate suppliers and manufacturers on the standards and processes for creating high-quality 3D models. This helped streamline content creation and ensure consistency across their product catalog furthermore they joined the Khronos Group, contributing to the 3D Commerce Working Group.

Porsche

Porsche is also actively exploring 3D technologies and building a robust infrastructure to support its 3D artists. Their process begins with XRite scanners for sample scanning and utilizes Autodesk products for rendering and asset organization. Below is an overview of their infrastructure. The original source of the image, along with a detailed video explaining Porsche's technical setup, can be found here: https://boards.autodesk.com/automotive-innovation-forum/design-studio-4f0f/porsche-create-manage--visualize

IKEA

IKEA has replaced traditional photography with 3D-rendered images for its product catalogs, significantly reducing costs and increasing efficiency in creating photorealistic visuals. Their journey started back in 2006, when they created the first CG product in a catalog and continued until 2015 when they launched a large scale 3D production and continues until this very day.

You can find more details and the source of the image below here: https://www.youtube.com/watch?v=Kup0d4Te3n0

Otto

OTTO has expanded its CGI expertise, investing €2.6 million in an internal CGI database, and creating seven new positions in this field. In the medium term, the team will grow further, aiming to represent at least 70% of the entire furniture range and 35% of all home and textile products using CGI by the end of the 2021/22 fiscal year.

You can watch more details about the approach here: https://www.youtube.com/watch?v=Kjh_KDmfG4g

Issues of Scaling

Through our extensive experience collaborating with companies trying to scale their 3D production, we’ve identified a recurring misconception: many believe the primary obstacle to scaling 3D is the inability to create 3D models quickly enough. While this might seem like the main challenge at first glance, as soon as you dig deeper this turns out to be very far from the truth.

Consider this analogy: imagine setting up a factory to manage 10,000 Excel sheets, with plans for 100 people to simultaneously work on them in the future. At first, creating the initial batch of Excel sheets might seem daunting. However, while it’s no small effort, it’s a straightforward task. Hiring enough people to generate the initial content doesn’t require groundbreaking innovation.

The real challenge emerges when 100 people need to work on these sheets simultaneously. Without cloud-based systems, this quickly becomes unmanageable. You’d have to precisely track who is editing what, enforce strict file-naming conventions, and constantly manage synchronization, yet even these efforts would probably lead to some level of chaos making it really hard to scale any further. On the other hand, using a cloud-based system inherently solves these problems, making scalability achievable.

To follow this analogy, some companies then take the approach that they would just rely on the old legacy formats, and develop the synchronization and update mechanism by themselves. This approach, while it might work in specific cases, is just as resource-intensive and wasteful as building an Excel sheet synchronizer instead of leveraging tools like Google Sheets or Microsoft 365.

Another often-overlooked factor is that the creation of 3D models is actually quite standardized. There are common formats, and nearly all 3D artists start by learning 3D modeling, as this area is supported by a lot of training materials. 3D materials on the other hand are a lot more difficult to create, and if you take into account that you would like the materials to be reusable everywhere (not just to make them work in a specific scene from a specific angle).

So in the end it turns out that the creation of the 3D models is the only easily scalable part of a companies 3D effort. Everything else is a lot more difficult: having the underlying cloud based 3D system and standardized surface digitization technology is a lot more difficult to achieve.

Solution

Over the years this is the main reason why we dedicated most of our efforts into these two areas:

Because we think this is the reason that most of the companies get stuck at scaling their 3D production.

Our 3D system is built around a Single Asset System concept, that allows you to use the same source 3D asset for any application. Combined with its cloud-based architecture, this design enables seamless scaling, even for teams of 100 or 200 people working simultaneously on the same set of 3D content.

Single Asset System

Our scanner digitizes the true reflection (svBRDF) properties of surfaces making it possible to use any material that was scanned without modification in any scene from any angle.

For detailed information about the technical setup and usage of our system, please refer to the documentation.

Typically we would list our competitors here as well, but at the moment we are not aware of any other solution that provides the same 1) native 3D cloud functionalities and that maintains 2) photorealism. As soon as you take away photorealism (that simply provide a way to upload 3D assets in a limited way) or if you take away the cloud based aspect (that simply work on your PC) there are of course many other solutions out there, but we have not listed these solutions here. Nevertheless if you are aware of any solution like that please send an email to info@colormass.com.