A) 19″ Living Rack

Project developed by Léa Pereyre

19” Living Rack is an open source variation around the standardized 19” computer server rack (or cabinet). Dedicated to be distributed in domestic environments for personal or for small size community uses, the modular base of the standard rack is getting combined with additional functions, to address this renewed small office/home office context.

In making clear reference to the famous Ch. Eames toy, the modular House of Cards project, the 19” Living Rack comes in one technical “Base” that can then be combined and customized into three different types: “Office”, “Home”, and “Garden”, to set up personal and somehow undersized data centers.

For each configuration, air flows have been taken into consideration and act as design and functional factors: the air in the front part of the rack remains temperate before entering the rack and cooling the servers, while the back and top air flows are getting warmed up and dried due to the computers heating process. “Office” functions comes therefore mainly in the front part of the rack, “Home” on the back where elements can be tempered or dried, while “Garden” comes on the top, equipped with moistened plants to clean, re-humidify, perfume and cool down the air.

A “Home Cloud Kit” (evolution)

Note: the purpose of a “Home Cloud Kit” (working title) has been described in a previous post. It will be composed by four artifacts which will become the main outcomes of the design research Inhabiting and Interfacing the Cloud(s), along with one book about the ethnographic field study and another one about the design research process.

Below are four links leading to four posts describing and analyzing the current state of evolution for each part of this kit. We expect the research and the “kit” to be finished by the end of March 17.

The “kit” will be distributed freely at the end of the project.


The final phase of our research consists in the prototyping of artifacts which relevance have been identified along the process. Tools, infrastructures and services are therefore addressed and will constitute a “Home Cloud Kit”.


This final phase is organized into the four following lines of work:


A) A Personal Data Center (evolution, models)



B) I&IC’s OwnCloud Core Processing Library (evolution)



C) A Personal Cloud (evolution)


D) My Data Controllers (evolution, models)


A Personal Data Center (evolution, models)

Note: “A Personal Data Center” (working title) is part of a home cloud kit, which was described in a previous post and that will be composed by four various artifacts, both physical and digital.

The kit will be distributed freely at the end of the project.


After a few design iterations through sketches and a bit of 3D modelling, we recently produced a set of first prototypes of what our domestic 19″ server rack could look like and how it could handle domestic functions as well. As a matter of facts, we can consider this work an alternative approach to what was set up and analyzed at the beginning of our research, when we assembled our own “(small size) personal cloud infrastructure“.

Our approach was fueled by several references, the first one being House of Cards, by Ray and Charles Eames :





The modular, simple and intuitive assembly process guided us for its adequacy within a Do-It-Yourself user context.

“A Personal Cloud”: a home cloud kit for personal data (centers) / “reappropriate your dataself”!

We’re entering the final straight of the research project Inhabiting and Interfacing the Cloud(s) and we can give at this point a first glimpse of the four design artifacts we are working on at the moment. They will constitute the main outcomes of our joint experimental effort (ECAL, HEAD, EPFL-ECAL Lab) and a kind of “personal cloud kit” (explained below). These creations will be accompanied by two books: one will present the results of the ethnographic research about “the cloud”, the other will present the design research process and its results – both in pod/pdf.

We already pointed out in the recent post “Updated Design Scenario” where we were heading. Since then, the different projects were better identified and started to get shaped. Some got eliminated. Prototyping and further technical tests are running in parallel at the moment.



From the original “final scenario” sketch to …


… a “Personal Cloud Kit”, composed of various physical and digital modular artifacts.


What emerged reinforced from the main design scenario is that we seek to deliver four artifacts (some physical, some digital, some combined) which themselves will constitute the building blocks of what we’ll call “A Personal Cloud Kit”. All four parts of this kit will be openly accessible on a dedicated website (e.g. in a similar way to what OpenDesk is doing).

The purpose of this “home kit” is to empower designers, makers and citizens at large who would be interested to start develop their own cloud projects, manage or interact with their data or even to set up small scale personal data centers at their places (homes, offices, garages …)

From design research wrap-up to final artifacts, updated design scenario (in scribble mode, #2)


This post consists in an important update to the previous note “From design research wrap-up to final artifacts, a design scenario (in scribble mode, #1)“.

It’s main purpose is to narrow down the previously sketched scenario and become more precise about the possible artifacts we will develop. In doing so, the present description shows a likely path and tries to keep some coherence within the overall design that is segmented in four different areas, often combined (software, hardware furniture, responsive objects, visualization). Yet and even so several objects and functions are described and named in this post, it will continue to serve only as a general blueprint for the last phase of our I&IC joint design research, while the final outputs could still largely evolve, based on the same ideas and plan.


These ideas keep their importance though:

Based on the graphic “Motivations <-> Usages <-> Problems” and its description of procedures, based on our “Design Learnings” too, we’ve tried to translate and objectify these into “natural language” of actions (and problems). At this stage, this is still a trial, but we’ve listed five pairs of words that work in opposition (verbs vs. past participles) and that cover the spectrum of functions and procedures: To Care (vs. Neglected!), To Accumulate (vs. Vanished!), To Multiply (vs. Shrinked! ), To Freeze (vs. “Melted!”), To “Pimp” (vs. “Jumbled!”) — all these corresponding to a set of cloud actions and explained with more details below, after the break —

These 5 pairs of terms will drive the development of an alternative, domestic and hopefully objective Cloud (“Our Cloud”), built upon the open source software OwnCloud with the help of our own I&IC OwnCloud Core Processing Library (which will be further edited and editable therefore). This personal Cloud will have the opportunity to be hosted into a new type of diy (and domestic as well) 19″ Cabinet.

The 5 pairs of terms will further drive the implementation of 5 Controllers or Network Data/Bot Objects (“Smart Objects”). The aim of these “controllers” will be to give an everyday physical presence to each user’s personal Cloud, to its 5 main folders, contained files or data and the processes they undergo. The manipulation of some of these objects will include the idea of “natural interface” or “gestures”.

In addition to these 5 physical controllers, 1 or 2 “root objects” could help monitor and possibly moderate the overall behavior of “Our Cloud”.

For the development of these “smart objects”, we’ve decide to take into account the kind of objects or infrastructure that are already present in a domestic environment, things like single functional objects (i.e. lamps, electric plugs, consoles, mirrors, clocks, etc.) and revisit them with a slight sculptural approach. We’ve also decided to consider a “language” that takes into account “invisibility” or “immateriality” (i.e. electricity, electrostatics, magnets, light, reflections, air currents or temperature, dust).

Finally, all of the above should be distributed as open source.

Towards applied object design and interaction

As we are now well into the second phase of Inhabiting & interfacing the Cloud(s), our aim is to materialize and verify our research. The intent is to work from two complementary perspectives: object design and interaction design. Lea Pereyre, object designer, and myself Lucien Langton, interaction designer and both assistants on the design research I&IC, will work on some parts of the physical hardware of the cloud as well as on it’s functions and usages.

Doing so, we are pursuing by design means the Learnings from the “I&IC design research wrap-up of sketches, towards artifacts” and “blabla”.

If the cloud as a medium has not yet been colonized by designers, expect of course in the applied case of user experience and interaction design practices, it is perhaps because it takes the place of a “blind spot” in our lives as front-end users. As soon as we make a gesture towards it’s nature it seems to vanish in a blur.

A first step in enabling designers to grasp the concept seems indeed to setup an enumeration of the building blocks composing the cloud. After all, it is a system of systems, optimized to perform certain tasks: upload, download, synchronize, share, compute, stream, and perhaps more. Many of them on third party hardware. The term “Cloud” is only a packaging for such terms, therefore it seems evident to investigate the possibility of a plurality of objects in order to create a relevant design for the somehow blurry term.

Moreover, the most flagrant incarnation of the cloud resides in it’s physical infrastructure. Colossal amounts of servers, electricity and cables are necessary to maintain this invisible yet crucial part of our contemporary society. This ecosystem built for machines to host and compute data is filled of objects, engineered in the never-ending quest for optimization: hubs, ventilators, connectors, etc. The most recognizable and useful object of this data paraphernalia is the 19″ server-rack. It is therefore natural to identify the server-rack as an obvious subject of product design in this research, especially also because it was “unblackboxed” almost at the beginning of this research.


The 19″ Server Rack and the “U” unit: object design

Historically, the invention of the 19″ server rack is a mystery. It was first standardized by Bell Telephone Company (later  AT&T), but was probably introduced beforehand in the railway infrastructure. Anyhow, it has always been an object dependent of the technology it supports in it’s evolution. This strange object has constantly evolved through the quest for performance optimization in electronics but it was until recently never questioned in it’s mobility aspects, let alone in itself as a designed object. The cloud’s physical infrastructure is a hostile environment for humans, even though it is meant to provide end-users with an abundant set of resources.

This hostile hardware is defined by a peculiarly standardized characteristic:
- The unit U  is specifically used to standardize server-rack heights and inner vertical spacings. One U is equivalent to 4.445 cm, or 1.75 inches.

This unit is of specific interest for our research as a formal constraint in object design. Other standards of interest were already introduced and documented in a previous post here.

It is necessary to acknowledge that these standards are maintained in order to keep in place a technological (and therefore economic) compatibility and interoperability within the infrastructure. It is difficult for non-technical users to gain access to the hardware necessary to setup their own infrastructure, precisely because the technical aspects aren’t communicated to encourage public use. This is why we need to re-appropriate these standards as designers with an open-source approach in mind. The first step towards this re-appropriation is to get familiar with the technical aspects of server-rack.

Considering the personal cloud as a paradigm serves primarily the need for users to store their “memory” (storage is by far the most popular function of the personal cloud), it is crucial to notice that by externalizing our “memory” to the cloud we have actually  displaced it to distant data centers. In this regard, re-appropriation is also a way to become conscious of this phenomenon.

Two references (out of the datacenter field) were identified as interesting design tracks according to our research questions (didn’t we wonder in this document about inhabitable shelter/data center in the chapter “What is the most appropriate approach?”)

The first reference, then, is Living Structures, Ken Isaacs, 1974, in which the author investigates our daily domestic micro-environments and synthesizes modular counter-proposals under the form of an open-source manual. The second comes from Shaker furniture, in which the aim was to design a universe of objects which all share a common function: to optimize the surroundings in order to liberate space for spirituality.


ken isaacs chair


Ken Isaacs, Super chair (top) and Shaker chairs (bottom).


The Functions: interaction design

As users, we often use the cloud without even knowing it. The reason for this is that the cloud is a system engineered to assure a constant access to data and other users regardless of their position on the planet. This goal to access everything whenever and from everywhere relies on key functions which are kept hidden (“blackboxed”) in the user’s experience. Therefore, our clouds, phones, computers and connected devices constantly upload, download, synchronize, compute, stream and share data in the background.

The closest way in which these functions exist for the users are in the form of buttons, icons & notifications in the user’s interface. We can see these actions are always triggered by the same user interactions: click, tap, swipe down to sync, scroll down to load more data (in essence download), etc. On one hand the user has an extremely restrained contact with these actions, but on the other hand an ever-increasing universe of connected objects embodies the granular appliances of these actions.

These connected objects, often referenced to as the “Internet of Things”, decline cloud functions under every form. These remain nevertheless opaque to the user’s eyes, which is once again a flagrant proof of the vacuous engagement in the design process to produce these objects.

Connected devices of the internet of things are often the only ambassadors of the cloud under tangible form. However, one cannot help to notice their aesthetic and interactive aspects are often disappointing. Indeed, these objects are only designed with the purpose to add market value to a technology, but very rarely does it question or reduce the system’s technical opacity. We believe designers have a strong responsibility in this.



Iotlist entered as a search term in Google images gives a quick glimpse of the trend in connected object design.


Designing a family of connected objects seems like a good lead at this stage, because it resolves several conceptual problems. First of all, it enables to kickstart a debate on the bundled nature of cloud functions. Dissecting the cloud into a list of functions is already a first step towards the establishment of an honest familiarity with the concept. In the second place, we need to give the cloud a body. One of the major difficulty with cloud computing is it’s invisibility. By embodying these functions we give place to commentary on each one of them.

Finally, objects are interesting in their different sizes and physicality. What would happen if instead of launching an upload with a click, it would be launched by temperature, location or proximity? What would happen if the object could be transported in a pocket? Or on the inverse, would be too heavy to move but very fragile? These are the tracks we are digging and interrogating at this stage in the research, enabling us to envision a personal family of objects related to the cloud.