
MVRDV.Cloud
In the age of overwhelming digital production, how do you retrieve and organise the masses of data that get discarded into an ocean of information? To make MVRDV’s archive comprehensible and approachable to a public audience, MVRDV NEXT created MVRDV.Cloud, a repository that doesn’t just focus on finalised architectural designs, but also presents information from the design process in a way that can be sorted, explored, and understood.
- Location
- Rotterdam, Netherlands
- Status
- Design
- Year
- 2021
- Client
- Het Nieuwe Instituut
- Programmes
- Exhibition
MVRDV.Cloud was not only designed to efficiently organise, process, and make sense of extensive and intricate digital data, it was also envisioned as a platform that ultimately translates this complex information into a format that is both accessible and comprehensible on a human level. It creates a large interactive data set that user can zoom into and out of, revealing images that can be sorted by theme, by their relationship to one another, and filtered by attributes such as size, colour, project number and more.
Initially developed as part of the MVRDVHNI exhibition, MVRDV.Cloud uses computer vision and metadata analysis to explore the contents of MVRDV's digital archive. The prototype serves as an interactive, self-learning map that organises and displays the complete design production of MVRDV’s first 400 projects, using data that was donated by MVRDV to the Nieuwe Instituut’s archives in 2015 – the first primarily digital archive in the museum’s collection.
In addition to its function as a repository, the Cloud also serves as a vital tool for preserving and safeguarding the legacy of MVRDV. It provides a comprehensive view of entire design processes, allowing for a level of detail far beyond that found in traditional project presentations. With its ability to visualise large digital archives in a systematic and ordered way, it allows for forgotten design concepts to resurface, leading to wide-ranging possibilities for exploration and new opportunities for reuse and inspiration.
MVRDV.Cloud’s strong filtering capability allowed visitors of the exhibition to navigate the large visual database with ease through its custom generative AI models. However, while the AI is capable of making intelligent connections, it is not without its flaws. Users are therefore invited to tag images themselves, directly contributing to the improvement and customisation of the AI's generative models. The interactive approach of MVRDV.Cloud allows for a more refined and accurate categorisation, while simultaneously training the system to better understand the nuances of MVRDV's design language.
Gallery
MVRDV.Cloud on display at MVRDVHNI
MVRDV.Cloud on display at MVRDVHNI
MVRDV.Cloud on display at MVRDVHNI
Screenshot of MVRDV.Cloud
HNI collage 02
HNI collage 03
Screenshot of MVRDV.Cloud
Screenshot of MVRDV.Cloud
Screenshot of MVRDV.Cloud
Screenshot of MVRDV.Cloud
Screenshot of MVRDV.Cloud
Screenshot of MVRDV.Cloud
Screenshot of MVRDV.Cloud
Screenshot of MVRDV.Cloud
Screenshot of MVRDV.Cloud
Screenshot of MVRDV.Cloud
Screenshot of MVRDV.Cloud
Screenshot of MVRDV.Cloud
Screenshot of MVRDV.Cloud
Screenshot of MVRDV.Cloud
Screenshot of MVRDV.Cloud
image























Credits
- Architects
- Design Team