Hello design enthusiasts and welcome to the design revolution newsletter.

Our goal is to motivate everyone with this free newsletter and be part of the design revolution by embracing AI tools!

We want to be as helpful as possible, so please email us with questions or feedback, and especially designs you have created!

This is what we will be discussing today:

• The advancements of real time 3D scanning and rendering in the AEC industry

• The advantage to 3D Gaussian Splatting

Here we go!

TODAY’S PERSPECTIVE

3D Scanning and Rendering in the AEC Industry

Until recently, 3D scanning has been largely dedicated to individuals or companies that have laser scanners or use a technique called Photogrammetry, where a series of photographs from different angles and specialized software stitches those photos together and creates a 3d virtual environment.

From Photogrammetry to NeRFS to 3D Gaussian Splatting:

A research paper published in August 2020, “NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis”, demonstrated a novel way to synthesize new views of complex scenes with unprecedented levels of detail and photorealism. This marked the inception of NeRFs as a significant advancement in the field of computer vision and graphics that uses a machine learning-based technique. However, even with this advancement, 3D scanning was still reserved to those with advanced hardware to synthesize large amounts of data into a cohesive 3d virtual environment. 

Almost three years later, 3D Gaussian Splatting changed that. Even though Gaussian splatting was discovered by the medical community in 1988, 3D Gaussian Splatting in 2023 was introduced as a new method to render 3D scenes in real-time. Similar to NeRFs, it’s still a radiance field, a model used to represent how light interacts with surfaces and volumes in a scene, however the difference is 3D Gaussian Splatting processes images exponentially faster than NeRFs and does not require advanced hardware.

Anyone can use 3D Gaussian Splatting:

3D Gaussian Splatting is accessible to anyone, either by heading over to websites like Polycam  or installing it locally on your own computer, this way is a little more involved but free, check out The NeRF Guru.

Use cases for this advancement have yet to be tested in professional environments, mainly because it’s a new concept and, similar to AI, AEC firms are still debating on the cost and exposure to introducing new techniques into their existing workflow.

The advantage of being able to 3D capture and process environments in real-time, with a minimal amount of hardware requirements, is opening the door for startups to offer new types of services or even smaller companies to gain more traction in the industry. And with 3D Gaussian Splatting only two years behind NeRFs, the advancement of computer vision and graphics has been accelerated and questions where we’ll be in the very near future.

That’s all for now!

During our reboot, we’re creating new videos that will help inform you about AI. Please stay tuned! Feedback and questions are much appreciated - respond here or send a quick message to [email protected]

Stay Curious and Watch This Space!

Warmly,

The Page Turn Studio Team