Develop electronics for us and the profession, and a hobby, and a way of self-expression.
in free time at night and on weekends, we are working on internal projects, one of which has recently become our first own Commercial Product. Here’s how it was:
→ First post here .
Once, an unknown person hacked into the Yulmart account of our CEO and bought a USB flash drive and a package of diapers. After this blatant incident, we revised the data protection policy, changed the passwords for work and personal accounts, and implemented KeePass to store accounts. This is a universal open source software, ready to use out of the box. But there is no limit to perfection: if you can improve something, do it.
During the brainstorming sessions, the idea of Pastilde, an open hardware password manager compatible with KeePass, was worked out. There was a strong feeling that the topic was not interesting to us alone. For the work!
The idea will not become a device without a plan. To evaluate the project, we divide it into small hw- and sw- tasks. Apparatuses and programmers estimate the laboriousness of each and contribute the expected values to Redmine .
The figures are approved by the managers, after which we usually prepare a proposal to the customer and a calendar plan for the project. For the calendar plan, we consider that
The coefficient is empirical, but it gives sufficient accuracy, since it is calculated on the basis of the plan / fact for past projects. To do this, each engineer takes his time every day in the same Redmine.
For an internal project, a preliminary assessment is no less important than for a commercial one. It brings us closer to an understanding of the real cost and timing of work, will give a chance to stop on time if the costs are excessive, and there is no result.
But here are hundreds of cups of tea and coffee behind, Pastilda revision 0.1 is collected, stitched and tested. Taking into account the emerging issues, we have compiled a list of the board's improvements (replace the controller, add a memory card, podshamanit slightly here and there) and software ( rewrite everything ).
We dreamed rather to launch Pastilde into the series and start shipping by hundreds of thousands, but it was necessary to study the demand. The prototype was shown to friends, they wrote on Habr , spoke at the conference , held presentations in several companies. The response to the device was positive: the problem of protecting accounts from hacking is familiar to many. But the dream of a major customer from Russia did not come true.
Then, to light up in an English-speaking environment and (if you're lucky) to cover at least part of the cost of the prototype, we went to crowdfunding.
Kickstarter and other pop platforms immediately dropped. The success of the campaign there directly depends on the budget for promotion, besides it requires 100% involvement of the team, which was impossible for us - along with Pastilde there were commercial projects. There was no experience in crowdfunding, so I wanted something warm, tube, and it is desirable to have it on a turnkey basis.
Leafing through the internet came across the Crowd Supply platform. At a glance at the main page it became clear: here it is, a means of collecting American people's money for Russian custom electronics.
It was striking that even small iron projects there were successfully funded, and the amount collected was often several times larger than the plan. In contrast to the same Kickstarter, 100% of funded projects come to the result (delivery of devices to bekers). The founders are proud of
this, and not without reason.
We were assured that the campaign would not be launched without a detailed study of the prototype by experts, and after the launch we would be provided with media support. It sounded good, we signed a contract and called a courier to deliver samples to Portland, Oregon. But then something went wrong.
During the filling of courier papers, it turned out that taking something from the homeland across the border, which contains the encryption function, even on standard components and well-known algorithms, is not easy. And if successful, we had to do it repeatedly. Frustrated, we let the courier go and sent the package by Russian Post. It was a ride, but with the delivery of the party it was necessary to decide something.
Help came from new foreign partners. They were taken on their own to organize the Pastild shipping company from the USA, and they offered to place an order for production at a local small factory , where a 10% discount is provided for Crowd Supply customers.
We did so, since the price was even lower than in St. Petersburg. The calculation of the price they have is automatic, you just need to download the source of the board and select the necessary components. Very convenient, we will come to this, with God's help, after about 30 years.
And to deliver ready-made boards around the world from the USA came out cheaper. There are more transport companies there, prices are lower, market hands, demand curves, that's all. In addition, we reduced the costs of moving the collected money across the border. Four birds with one stone, yuh!
Something, however, slightly upset. Media support turned out to be more formal: publications appeared, but monotonous and not for long. Comments remained unanswered. We didn’t wait for feedback from the experts, and later we found out that none of our 5 samples were ever included. Okaay.
The secret of overfulfilling fundraising plans turned out to be simple. On the advice of the curator, we lowered the initial level nine times , eventually exceeded the plan by half =)
The firmware cost $ 1 per minute, which is inexpensive by local standards: the hourly rate of an experienced electronics engineer in the States is $ 150-250 per hour or even more.
Of the 324 boards, 30 did not turn on: the loader worked incorrectly. We quickly resolved the problem, but had to pay for the firmware again.
The project now and then resembled a group rake run, but we made good progress. The other day, Crowd Supply began sending out ready-made devices. In addition to Russia and the United States, Pastilde is expected in Germany, Italy, Canada, Australia, Singapore, Ireland, Norway, Switzerland, Finland, Britain, Slovenia, Israel, the Netherlands, Austria, South Africa, Denmark, Romania, New Zealand, Japan, France, Belgium , Spain, Malaysia and the Czech Republic. When the parcels fall into the hands of recipients and the most interesting begins. The software development department is already in anticipation of fun:
There is still a chance to order
Pastilde, we made them with a small margin. You can also take the source code on the BitBack
to build and flash the device yourself. Redo it for yourself as it pleases, we are not greedy, this is open hardware.
Few digits last:
The complexity of the project (as of 01.10):
Initial campaign target: $ 50,000
Adjusted Target: $ 5,500
Start campaign 02.02
, the goal is reached 04.04
(2 months and 2 days)
Total collected $ 11,975
(217%) from 191
Start of production: 08/10/2017
In parallel made commercial projects: 8
I have it all, thanks to those who believed in us!
Take a look at some of our other designs, internal and custom .
With approvals and criticism please in the comments, on cooperation - in a personal. Put likes, use complex passwords.