Continuing my real-time visualization exploits, I decided to bring one of my past visualization ideas to completion. The purpose of this visualization, focusing upon the 2017 model-year Chevrolet Volt (2LZ trim), was to idealize and future-ize either end of the product development spectrum; from the initial product design stages to the final customer-facing marketing materials and other point-of-sale materials (POSM). The idea of the virtual design studio as well as the virtual showroom are compelling ones that not only continue the modernization of the overall design process but also provides a logical progression of how the potential end-users may come to expect to research, purchase, and interact with the products of tomorrow.
Focusing on the idea of the virtual showroom, this visualization idealizes a situation in which the user navigates to the product manufacturer’s website (in this case Chevrolet) on some device and picks a vehicle that they would like to learn more about, and chiefly — visualize it in various states of trim, functionality, and color. The user would essentially have a photo-realistic, computer-generated, build-and-price experience that they could access (the same, in practice, would apply to designers viewing the various stages of their design process).
Particularly important for maintaining the iterative nature of the design process, hardware advances (CPU, GPU, etc.) as well as software advances (in this example, Epic Games’ Unreal Engine 4) allow relatively-massive amounts of data to be viewed in such a photo-realistic way. This dataset (the CAD model of the car itself), while it does not contain an interior in this visualization (seats, dash, center console, etc.) contains upwards of ~5 million polygons on-screen at any given point (for the entire vehicle on screen). This means that designs/products can be viewed (if at least only locally) much more rapidly and without as much decimation or reduction in the amount of data being viewed (as large datasets would typically be difficult to setup and ultimately render at any usable frame-rate).
As a result, such a large dataset can be viewed at upwards of 60 frames-per-second at HD resolutions with photo-realistic lighting and other post-process-effects (bloom, depth-of-field, bokeh, etc.). The user has complete orbit and zoom control of the camera/view. The user can also switch between several pre-determined paint options. The specifications and capabilities of the local computer running such a visualization is key to how well such a visualization performs; I will include the relevant specifications below (this is just another one of my bespoke systems).
Software used for this visualization:
Unreal Engine 4
Autodesk Maya 2016
Foundry MODO 701
Intel Core i7-6700k
NVIDIA GeForce GTX 1080
Samsung NVMe SSD
16GB DDR4 RAM
Windows 10 64bit
Winston Patrick Brathwaite