Une série d’articles relatant l’historique du projet, dans l’ordre chronologique pour une lecture plus naturelle :)
The idea of creating a free smart home brain didn’t come out of thin air.
I had been looking around every now and then for a way to build a smart home system for several years, as part of a (unfortunately still) future project to build myself a house. In the course of 2010, people also started asking me advice on the different existing solutions to this smart home technologies dilemma I was still pondering.
There are many options to create an autonomous system: cheap ARM boards, from about a hundred Euros for the simplest models with a few hundred MB memory and a few MHz to recent beagleboardTM or pandaboardTM, other similar boards, plug computers, and up to “sticks” based on the Raspberry PiTM model. But all these solutions come with new problems: too few connections, no box, no screen, no handling of smart home protocols, expensive, and usually more than one of these. And I wasn’t impressed with the alternative of using a PC (expensive, noisy, power hungry) with a USB key per protocol (expensive as well) either.
A few month before, I had bought a (well, 10 actually) plug computer and a USB touch screen. This seemed like a good starting point, especially since I had been able to easily run my favourite Linux distribution on it, the one I was planning to use as a basis for my smart home platform.
But problems crept up soon enough.
Out of the ten plug computers I originally bought (bulk buy to save on P&P and import duty), two died out really quick and the others suffer from serious heat dissipation issues. Specifically, they restart every time they overheat, and they overheat every time both their Gigabyte Ethernet interfaces are used…
Moreover, connecting the screen through USB uses up both USB ports… no more room for a USB key, a hard drive or a smart home USB key.
This is when I started to envision a custom made aluminium case. I happened to know someone (the creator of Langear) who made compact PC boxes. We could have integrated in one of those neat cases a plug computer, heat dissipator, a USB hub to get a few extra USB ports and a screen modified to use only one USB port and be powered by the internal power suply.
After some pondering about the general specifications, we hit a wall: the availability of parts to integrate. The selected touch screen was already out of production. This had already happened to me a few months before on another project, stopping it dead in its tracks. I wasn’t getting myself in a situation like that again.
Around that time, I was making a sizeable contribution to the creation of a system for a client and I started to realize that creating a product from scratch wasn’t as daunting as I thought.
This prompted me to start the project (then called DomoPlug) in August 2010 to create a product on my own terms instead of having to work on creating some industrialist’s product. This allowed me to design an open, community focused product, an give free smart home technologies a breath of fresh air.
An efficient team
The three people who led the DomoTab project from the start are all passionate about smart home technologies and each have over ten years of experience of managing technical project and creating companies.
Working in a group tends to make things easier, especially since our skill sets complement each other:
- People management, quality assurance, leading international projects and conception and development of medical systems
- Embedded systems development, knowledge of both informatics (Linux kernels and embedded Linux systems) and electronics (micro-processor and micro-controller systems)
- Leading Software as a Service projects with both start-ups and established companies
This allowed us to start on our own terms, outline specifications, list steps, study doability, start work on a business plan, a company plan and overall move forward in spite of disagreements that we were able to solve through dialogue. No big deal when the ambiance is good.
The one problem was that we were only moving forward virtually…
The next step, once all that was committed to paper, was to try to replicate it in reality, but to do that, you need “paper”. One of us had had experience of working with a business incubator and we decided to follow that path. Especially since a R&D and innovation focused project like ours was entitled to grants.
Thanks to my contacts at the école supérieure CPE-Lyon, we easily found someone with whom to negotiate a partnership with the Electronics, telecommunications and embedded informatics laboratory at CPE-Lyon, this partnership being a requirement of the incubator we selected.
The DomoTab project officialy started incubation following a commitment committee on December 21st 2010.
The business incubator offers custom, personalized support to start up creators for the length of their project’s development. They foster two types of companies: those monetizing an innovation developed in a laboratory and those collaborating with a laboratory to develop an innovation (the lab has to be specified at the start). DomoTab is in this second category.
This step gave us access to the incubator’s help including the training and presentations it offers, preparatory studies on commercial and intellectual property prospects, use of the lab but also a grant from the Rhône Alpes Region matching the lab’s investment up to a 30.5K€ ceiling. The lab’s investment will be (generously) paid back if the project becomes profitable but nothing is due in case of failure. The Region’s grant doesn’t have to be paid back but it is only awarded to the lab to cover equipment purchase and not to the project developers themselves. Any equipment purchased through this grant remains the lab’s property.
The study begins
All this gave us a platform on which to test our selected processor (TI’s AM3894) and enough equipment to evaluate various smart home technologies.
We were then able to go further and deeper in the technical study, fine tune specifications and conduct a good deal of research on smart home technologies and the protocols they use. We received help from two interns (Gabriel and Xavier) who worked on various protocol tests and implemented them on the processor test platform.
After extensive negotiating, we managed to get the incubator to finance an external doability study carried out by Adison, in order to compare it to the one I had conducted on my own.
Business incubation also imposes a number of less engaging tasks involving project monitoring and especially around the company creation aspect: presenting the project and its progression, brainstorming about the company to be created, how to outsource the aspects we couldn’t do ourselves, working out the respective roles of the developers in the future company, taking part in an “innovative company creation” competition, work on the business plan, work on positioning… and this is where things went wrong.
Side effects of incubation
Business incubation entails spending a lot of time (and then some) on non technical aspects of the project. This makes perfect sense for a project developed by people without any experience of company creation or who hardly ever made a presentation. However, we had created or taken part in the creation of 6 companies out which of 6 were still (and are still) in operation (Yep, that’s 100%).
Understandably, the incubator wants to check our abilities and the coherence of our project. It’s actually reassuring to discuss and justify our positioning and choices with them. It helps you go forward and validate or reject choices. The first time. By the third time, it feels like the objective isn’t so much to empower the developers as to format them according to “company creation 101″.
But our project aimed to innovate not only technologically but also by its openness. This implied a business model founded on principles from free software and social and supportive economy such as sharing and equality, and that was upsetting to some…
What came of that is that two of the three developers got exhausted and had to get back to their core activities for sheer survival. One can’t spend two years of full time work on a personal project, there has to be a source of income.
Of course there is another option. Finding financing or trying to sell the project to one or more people or companies, hoping they’ll content themselves with a minority share. But finding that kind of financing while keeping our open positioning wasn’t realistic, especially if we hoped to retain a majority share. We would have had to go from creator to employee.
We also tried to get government funding without success. Usual answer: the French government won’t invest in French innovation or in social and supportive economy. We couldn’t get help from the Chamber of Commerce or Oséo either.
So we went another way: crowd funding, an idea we had been pondering for some time
The first part of the financing came from our own family.
However, this type of financing is more adapted to purchasing equipment and services than to provide the developers with income.
In this end, it’s through the help of the Regional Union of Cooperative Companies (URSCOP) that we made a breakthrough at the end of 2011, leading to the creation of the Techno-Innov cooperative