As we did the first Russian smartphone, continued
So, in the courtyard on October 19, 2012, our trick with the prototype was a success and we need to urgently launch the development and production of a real smartphone. And by March, issue a batch of 400 pieces to the customer.
Summary of the first part
В принципе, можно и не читать первую часть, потому что вот её краткое содержание:
The problem is that in addition to the medal for the successfully held exhibition, we had little in the asset. That is, we perfectly understood everything that relates to plastic, design and mechanics, we designed several hundred cases for all sorts of different electronic devices, but at the same time almost did not understand anything in the development of electronics and smartphones in particular. And in general, it would be more correct to call them not smartphones, but data collection terminals, but we didn’t even know that then.
But we had a great desire and courage.
All past exploits with prototypes did not help at all, but rather even hindered - customers wanted to see exactly the same serial model, and any new proposals for design or layout were hacked to the ground. We had to convince them for a long time, even to remove the terrible "hump" with the battery from the back wall and insert the battery inside the case. The only advantage of this situation was that the design has already been blessed by the sky-dwellers of Russian Railways and there is no need to spend time “beautifully ugly” to argue.
The project's economy was built quite simply: to sell the first 400 units of devices to the customer at the horse price, laying all the development costs there. Then, on the series, this price can be dropped, leaving a reasonable marginality. The software was sold separately, I didn’t even go there. After all, we acted as a contract developer and manufacturer - and accordingly, we took a one hundred percent advance payment for all the works. Where did the integrator take this money, and how much he earned from it - these questions weakly touched us. As a result, we had a very simple business task - to develop and produce. All the difficulties of tenders, tests, guarantees and logistics fell on the fragile shoulders of the integrator.
Part two. Actually development.
Technical task for the development of software developers formulated, the company Aisa. On the basis of their hotelok, we developed a solution, and then the integrator wrote his TK, already at the Russian Railways tender.
The initial customer's wish list looked something like this:
- Display 4.5 inches, high brightness and contrast, to be seen in the bright sun.
- Touchscreen, which responds to pressing with gloves. No styluses - they will be lost.
- Chip family TI OMAP, in extreme cases, ST or Intel. No Media Libraries!
- Laser 1D barcode reader, optional - 2D.
- Shockproof and waterproof housing.
- Built-in Athena card reader for contact cards, connected via USB. Only on the native Atenovsky motherboard, because it works, and nothing can be touched there!
- Battery at least 5000 mAh hot-swappable without turning off the device.
- Fully open source files Android.
- No bluetooth and wifi - this is not safe!
- Connect to a computer via USB through the docking station and micro-usb connector.
With these requirements, I flew back to China to pick up suppliers. The simplest solution would be to find a company that has already made such devices and order them to do everything turnkey according to our requirements. If you fail to find such contractors, you will have to assemble everything from small pieces — the motherboard, the screen, the touch panel, the case, the battery, and a bunch of different small things that even you thought were scary, such as antennas or protective pads. To implement such a project without the support of a local Chinese partner is completely unrealistic, for no money.
We had such a partner - the Shenzhen company Gaodi Industrial with its plant and extensive experience in the production of mass electronics. Estimate our generosity: we immediately offered them to go to the project 50/50, while they are responsible for all Chinese contractors and make the final assembly at their plant - and we work with the Russian side and the rest of the world.
The option to "give someone everything under the key" was dropped pretty quickly. We found two companies that made similar devices, but firstly they had very weak chipsets, and secondly, they both asked for absolutely unaffordable development money - something around 600 thousand dollars.
With the second option, too, there were surprises. We did not find a single Chinese design house that worked with TI OMAP, ST, or Intel processors, but we did find several Qualcomm developers and an uncountable number on Mediatek. Push through Mediatek was unrealistic, so we launched a campaign for Qualcomm. This was not such a difficult task, firstly, because Qualcomm was already a market leader at that time. And secondly, around this time, TI announced that it would soon turn off the production of mobile chipsets, and the solutions from Intel and ST were quite exotic.
But even by Qualcomm, finding the right design house was not a trivial task. We looked at a dozen candidates. It is almost impossible to understand who is in business what it is. Some sit in a tiny little office with five of them, they say that only a month ago they worked for Huawei and asked for 120 thousand dollars for development. Others - two hundred developers in a huge office building, are already asking for 150 thousand, but at the same time it seems that the same people from Huawei outsource work. There were fashionable Chinese industrial designers ($ 300K), there was even a team in which at the same time was the development of the Yota Phone ($ 1.2M).
As a result, we chose one small company, which offered adequate prices for development and circulation, was ready to secretly give us Android sources from Qualcomm, and in the future to help with the transfer of production to Russia.
The body decided to give the contractor with extensive experience in the manufacture of secure devices. As soon as they entered their office, they saw a shop window, and immediately understood - our guys.
The truth then went to the production and a little sad. But they did not change their mind.
There was also a long search for displays, resistive touch panels, video modules and barcode scanners. It was necessary to hit the road a second time in Shenzhen - not so much to communicate with suppliers as to support the spirit of the Chinese team and show the seriousness of intentions. In parallel with this, Gaodi buyers collected the rest of the small things - there are about 20 suppliers, fifty different components: cables, screws and gaskets, speakers, microphones and waterproof membranes on them, various stickers, glass and nameplates - when I took on the project, then even I did not expect that there will be so much of this stuff.
In Moscow, too, work was in full swing: the layout, the updated design and construction of the smartphone were being made. We started with some fantasies about how this would be done and with each agreed component we refined our 3D model, getting closer and closer to full-fledged design documentation. At the final stage, engineers from the Chinese mold plant connected, and we have the final model!
This is a gypsum model, machined on the CNC - just hold it in your hands and understand the scale.
By December 20, when we received the first payment from the client (the client, as it should be, was wildly slowed down with payment) - we had almost everything ready: the components were selected, contracts with suppliers were signed, the 3D model was finalized. Two months of intensive work without money, without guarantees - for one honest word and huge interest.
As a result: the Catholic Christmas is in the courtyard, we have a lot of rubles in our account, and nobody canceled the final deadline for the supply of 400 devices in the middle of March. Have you ever started projects with China in late December? So you understand the scale of our problems. First, America rests for two days, and no payments go. Then Russia buzzes for two weeks and it is impossible to achieve anything from the customer. And in February, all the Chinese leave for the New Year holidays and then everything gets up. At the same time, we have only three months to develop electronics, make a mold, order components for the first batch and produce 400 pieces. And somehow deliver them to Moscow.
What is the development of a smartphone board? In principle, any chipset manufacturer grants developers a license for a reference design - an almost full-fledged smartphone, with all the documentation, software and drivers. Design House buys such a license (expensive, from $ 500K to $ 2M for one chip line) and sells solutions to manufacturers of smartphones at retail, modifying the documentation for a specific size of the smartphone or for tricky external peripherals. In our case, everything was unusual - and the electronics had to be placed on two boards and the number of modifications under the periphery was too high.
Judge for yourself: we need to connect two batteries, inventing charge-recharge and indication algorithms, we need to connect a resistive touch, three USB connectors, one internal - to the card reader and two external - under micro USB and a docking station, we also have a barcode reader UART, custom bright display and three mechanical shortcut keys functions. We had no illusions - it is impossible to do such work from the first time and without mistakes.
In principle, the participation of the Russian team in the development of circuitry was minimal - we received suggestions on various nodes from the Chinese, gave our comments and corrections - and things moved forward.
As a result, by some miracle, by February 5, the first engineering sample was ready. He even started, though he didn’t see the map and the barcode reader, but these are trifles. When you hold in your hands that which was recently only in pictures - emotions overflow, a complete feeling of a miracle accomplished.
All that happened afterwards was the usual production days. Tests, improvements, adjustments. Order parts for a batch, when you have not yet fully earned an engineering sample. Disputes with the Russian programmers, who even can not assemble the image of Android themselves and are surprised that it weighs more than 10 GB in the archive. The battle with the Chinese, who do not understand the obvious things and are in no hurry to share documentation.
Part Three: Production
Production unfolded in a clean room. There was no big sense in this, usually one clean workstation is enough to install the display and the touch panel. But the room was still empty and decided to take it in order to show the customer that everything is very difficult and expensive.
Test equipment for testing GSM radio signals is very expensive, you can only use Agilent or Rohde & Schwarz. Few people buy it - usually rented.
For functional tests made specialized equipment.
For all assembly operations, they made full-fledged technological documentation, everything as usual - nothing special.
And as usual, the first batch did not go smoothly - the touch panels did not work, or the GPS did not pass the test. As a result, the production of the first four hundred pieces took three weeks instead of the planned three days.
The integrator certainly swore badly, but there was nowhere to go - they accepted it and shipped it to RZD for field tests.
Then there were many months of trials and improvements in software and hardware. We reworked something there, replaced it, argued with the programmers and waited for the order of the promised series of 60 thousand pieces.
And in October, the next railway exhibition took place, at which our smartphone was noticed by a wide habopublic .
Unfortunately, the project did not go to the big series. We produced another 4,000 units in China, and then the management of Russian Railways began to change, and all such projects were permanently suspended in uncertainty. The planned transfer of production to Russia did not take place, and the integrator unsuccessfully tried for three more years to stir up a new Russian Railways team to something like that.
But you can not say that the story is over. Already in 2017, practically in the same composition, we made three models of online cash registers, and now we are developing five more interesting projects of android devices. So the most interesting is yet to come!