The journey all data has to travel to the Google servers, and in order to make sure their products work for you, is incredibly complex and involves a wide range of different technologies during different moments of it’s processing.

We spent many hours together with the Google Cloud team trying to recognize which key moments in the data journey were relevant to communicate in the experience, and the approximate scale and angle that each moment should have to convey all the relevant information in a simple and appealing way.

We decided to have 4 main levels of scale with additional graphics that could help us to explain special situations.

From the beginning it was clear that we recreate the infrastructure on a global level using visuals that could depict the whole planet Earth and the current Google backbone network, which includes both Earth and sea cables, Edge nodes, CDN PoPs, Edge PoPs and Regions.

The second scale level was Region. We made this using as many references from public images of the different Google data centers around the world. The main purpose of this level was to depict examples of important exterior elements that some Google data centers have, like cooling towers or electricity generators. We also wanted to show how a Google Data center could make smart and respectful use of the environment including electric windmills and farms near the centers.

The third scale level is the Zone level and it focuses on one of the buildings inside of a Region. We removed the ceiling to make visible all basic internal components, their possible position and the way they could interact with elements in the outside world. This model was one of the most difficult to create due to the real life complexity of data centers and the use of particular elements from the design of each building and their surroundings.

We switched to a front server cluster view in the fourth level of scale in order to depict moments related with data processing. We created 2D animations that used small blinking lights as a visual metaphor for the different kinds of data bits processed and shared inside the servers.

A last zoom level was where we depicted how to visualize a single server from a top view. The idea was to explain some of the most common chips and hardware used in Google server’s motherboards.

An especially challenging part was to create different visual systems to represent software elements, data states and other logical information that doesn’t have a physical shape or a universal depiction. Some of these elements were data queries, data in motion, kubernetes’s container systems, encryption keys, data chunks of data nodes.

The Google colors in motion were used as an extra element in the color palette to show the difference between static and dynamic data in different instances.

The use of a “logical view” was necessary to explain the different levels of encryption data wrapped by Google.

A condensed version of the "logical" view was also used for the Container Operations case to visually represent automatic scaling.