The making of a Cubbit Cell: a journey from conception to production

Art speaks to the artist, design speaks to the customer.

Finally, I had found the logo of Cubbit; the Cubbit Cell, though, the dedicated device that runs Cubbit’s distributed cloud — that was still a dream. Marco and Alessio were coding Enigma, Cubbit’s cryptographic engine, but there was no clarity about the Cell. Would it have to be a dedicated device? Yes — that at least was sure. There was no way the network would be stable on Raspberry Pi for a long period of time. Indeed, the ongoing beta test is based on a network of Raspberry Pis, but that is, you know, a beta test. The present is not the future.

For my part, I was attending university. I was not inexperienced anymore, but there was still a lot to learn. There always is. After all, the logo of Cubbit was my first, real work out of the university. Nonetheless, it was highly appreciated and gained me the trust of the team. Indeed, the founders believed in me and put me in charge of designing the Cell. It was a significant increase of responsibility: designing the logo is one thing, designing a functional object with electronics inside is something else entirely. Maybe not more burdensome, given the time and dedication I invested on the logo, but it was certainly different. It felt different. Probably, not even in the same ballpark. But let’s not rush things, shall we? If I am to tell the whole story, let me do it right.

Concept

It all started in 2016. At the time the logo was yet to come and there was little to no time to think: we were applying to a startup competition and desperately needed a prototype. Fast. Not of the software: what we needed was a hardware sample. A minimum viable product to apply and run the first user tests. A proof of concept, if you please. I was short on time. As you know, Cubbit is a cubby for your bits, but sounds like the English word “cube”, so the first thing that came to mind was — guess what — a cube. Some fantasy, uh? But believe me: we were literally out of time and I, as a young undergraduate, didn’t really know what I was getting into.

Prototype

Simple. Rude. Unashamedly ugly. So ugly it is funny to remember that I was not alone doing the work: there were, in fact, two other designers in Cubbit. Like me, they were students taking their first baby steps into the world of work. At that time I was probably lacking a work method that I could call my own. The other two designers and I were all on the same boat but, as a team of strangers rowing without coordination, we were getting nowhere.

A team is like a rowing-crew. A common destination is never enough if you don’t work together.

The cube we crafted had two major advantages: it was easy to build and, to a certain extent, consistent with the vision. Plus, it was impactful. Or at least, it was the most impactful thing we could build in two weeks. So we began. We armed with block notes, pencils, and 3D printers. For the prototype we made use of acrylonitrile butadiene styrene — ABS for friends. Connections, i.e. USB and ethernet ports, were shaped in polystyrene. I spent two days sorting out how to put polystyrene into the cube. Two. Entire. Days. Can you imagine? Not to mention the whole week I spent to understand how best to design the cube in 3D and to prototype the connections in polystyrene. In retrospect, this is really a story of many first times: it was my first time with a 3D printer, ABS, polystyrene and, basically, for the first time I was getting my hands dirty with a concrete project. I was 22.

Design-wise, I was still a virgin.

The result was three small cubes, each one built out of a different material: one made of faux wood, one white in raw ABS, and another one black. If you ask me why three different materials, the truth is we had two different 3D printers at our disposal, one belonging to a friend, another one to the university. Thanks to this, we could optimize the time and be prepared in case one of the two 3D moulds had gone wrong.

And then what?

All in all, the first prototypes turned out to be perfect for the job, as we managed to win a few startup competitions and survive. Soon afterwards, I graduated and got involved in many other side projects that, over time, helped me build and develop a personal, mature approach to design.

Mood-board

Although, my head was full of questions: how to design something that was at the same time easy to use and relatable to our original target at the same time? Was I supposed to entirely dismiss the cubic shape? Had it to stand out and get noticed or to blend into the context? But overall: how to convey the concept of the swarm through a single product? What’s more, I had to conceive the Cell from the get-go in order to make room for the ARM computer inside — which was, at the time, completely unknown to me. I looked at the problem. Thoroughly. Then I looked at it a little more, took pencil and paper, and tackled it with the tools of trade. First, a mood-board: I put everything on paper. And by everything I mean: sketches, unripe concepts, drafts — literally anything I could come up with to express our vision. What I was doing was, basically, a detective job, that is: plotting out any possible problem in advance, so to pave the way as smooth as possible.

The more you work in the present, the fewer problems you will have in the future. Future problems are always more time-consuming than problems of the now.

User testing

Once I had cleared any doubt about the ARM computer and the production process, I focused on the actual shape of the Cell. Two were the roads ahead: sticking to the cube or going for the hexagonal cell. I, for one, loved the hexagon’s idea from the guts, although

design is not rhabdomancy — it was not up to me to decide solely on the basis of my intuition.

Plus, intuition is worthless if the tests prove you wrong, and of users tests there was not a whiff. Not yet, at least. So I built not one, but two prototypes — one for each model I had in mind. I used cardboard first; that was a good idea, as it allowed me to immediately get a feedback about size, indentations as well as to understand how much space was left for the ARM computer inside. Afterwards, I 3D-printed a couple of models and prepared myself for a first public response. Indeed, we were headed towards the Maker Faire Rome 2018, where Cubbit attracted a lot of attention.

At the booth I gathered many feedbacks, interviewing the attendees and asking them what they were expecting from the sketches I showed them. Subsequently, I invited them to write down their opinions, so to better make sense of the wisdom of the crowd and establish a winner.

As expected, the cube didn’t win the heart of the Maker Faire audience, losing to its hexagonal brother, which got a staggering 75% of appreciations from hundreds of people. Also, the cell’s shape was much more coherent to the storytelling of the swarm. “Join the Swarm” already was our payoff. It was December 2017. I was relieved: many more user testing was to be made to confirm the choice, but I already knew I had found the shape of the Cubbit Cell’s embryonic form. Was it the end, then? Far from it.

Fine-tuning

Although tangible, the Cell we presented at the Maker Faire Rome 2018 was still vague and somehow fanciful. The inside space was unorganized, and a few brilliant feedbacks collected from the users’ tests were still to integrate. The hard disk’s positioning inside the Cell was a concern and the reason of many sleepless nights working. The distance between the ethernet port and the indentation was so thin that, once connected, the ethernet cable always got stuck into the port. Last but not least, the assembly inside the Cell was not as I expected, as the sketch was, to some extent, reinterpreted on grounds of production optimization, so that the indentations were often asymmetrical. Long story short, there were several issues. But I addressed them, one by one — first in the design lab, then working shoulder to shoulder with our hardware partner and with Eugenio, my colleague in charge of the production process. It wasn’t easy, as you could feel the constraints, design- and time-wise, imposed by manufacturing breathing down your neck. But it taught me that if you want to get things done and have it your way, you need to follow the whole thing from A to Z.

Design is like a baby: you don’t want to miss a thing.

And I didn’t, regardless of the difficulties.

More than anything, collaborating with the manufacturer forced me to backtrack, since many solution which sounded nice on paper turned out to be significantly more expensive or, simply put, inefficient. To pass the thermal test, for instance, I designed a circular air-hole around the Cell, which allows the air circulation and avoids overheating. Afterwards our hardware partner and I decided, in order to improve the performance and the internal stability, to implement a few air-holes at the base of the Cell as well. To further increase the hard disk’s lifetime I placed it inside the shell but below the ARM computer rather than on top of it, so to minimize the impact of vibrations. For the same reason I had air-holes put at the bottom of the Cell. The hardware partner suggested using screws, three hidden and one in plain sight, but users’ tests showed it worsened the UX, since screws implicitly tell you to unmount the Cell.

Production

Three years have passed since I armed with block notes, pencils, and 3D printers. One thousand, ninety-five days. Don’t they go by in a blink? The final chapter of this story is being written as we speak, and I just can’t wait to see the Cell in the hands of the users. This has always been our goal: to put the power in the hands of the users. I don’t know the future, but in my heart I know it is something worth all this time.