As a starting point, underground urban networks can be understood as the vital arteries of a complex system of flows that traverse and sustain the city — not only mobility, but also water, electricity, gas, data, and the regulation of urban infrastructure. The plastic pipes used in the Onsite project become, in this sense, a celebration of the city and its multiple channels.

By transforming these pipes into a piece of public furniture that brings people together, the project operates on the thin line between connection and separation — echoing the paradox at the heart of urban infrastructure.

A second vector relates to the use of computational design and digital fabrication tools. At the time the project was developed, a new digital architectural culture was emerging — driven by pioneering practices such as NOX (Lars Spuybroek), dECOi, UNStudio, and Foreign Office Architects, among others.

One of the central topics in the design discourse of the early 2000s was the notion of the non-standard — a paradigm shift away from repetition and modularity, toward continuous variation, customization, and the integration of form-generation with material logic and fabrication processes (see Bernard Cache, Greg Lynn, Kolarevic).

The Onsite proposal was thus conceived as a system capable of generating site-specific and customized variations, in direct response to the surrounding context and the needs of a given location or commission. In this sense, the project emerged from a design process shaped by both environmental parameters and computational logic, where form arises through adaptation rather than repetition.

As a result, the use of custom CNC laser-cut metal frames, combined with corrugated plastic pipes, offered a fabrication solution that required no molds and enabled adaptable, site-specific configurations.

The corrugated structure of the pipes gives the material two key properties: it can be bent along its length, and it is structurally strong under compression. These characteristics — combined with its light weight, its flexible nature, and the fact that it is an industrially mass-produced material — made it ideal for the project’s goals of adaptability, accessibility, and rapid deployment.

Regarding the design system developed within the Onsite project — what I referred to as the "programming of furniture" — a collection of typological profiles (a kind of functional alphabet) was used to define potential uses. Each profile corresponded to a specific spatial gesture or mode of occupation (sitting, leaning, reclining, gathering, etc.).

These profiles were then arranged along a guiding line and subjected to interpolation — or morphing — allowing the form to transition fluidly from one use to another, and translated through the piped fluid geometry that defines the project’s visual and structural language.

The OS.03 project, developed at Mediaruimte in Brussels, introduced an interactive graphic interface — a touchscreen on which visitors could draw piped profiles and generate variations of the system in real time. This participatory approach offered an early exploration of how users could engage with and shape design through digital means.

Today, tools such as Grasshopper and other parametric or procedural design platforms make it possible to quickly adapt and evolve forms based on input or context. Likewise, additive manufacturing technologies like 3D printing allow these digital models to be physically realized with increasing ease and precision.

What was once imagined as an ideal — a fully adaptable, user-driven, and digitally fabricated design process — is now becoming a tangible and practical reality.