Cloud for development companies: how Teslawatch smart bracelets use the cloud

We continue to talk about the use of the Microsoft cloud for solution development companies (ISVs). In this issue we present the story from the company Teslawatch, just recently released on Boomstarter with the second version of its smart bracelet.

Teslawatch was founded in the spring of 2014. The company specializes in the design, manufacture and sale of wearable electronics. Our first device was the Teslawatch T-Band, which we launched for sale in December 2014. In the process, we are developing mobile companion apps whose backend works in the Azure cloud.

Development tools

With development tools, everything is quite traditional - Xcode 7 , Android Studio , Visual Studio 2015 . We make the layout of the design with the help of the vector editor Sketch4 and the service of automatic construction of graphics for Zeplin devices.

Problems and how to solve them

We had some serious problems. The most serious is the possibility of losing the statistics collected by the user when reinstalling the application or replacing the phone. The second is the inability to work with the bracelet on multiple devices. These are problems with the lack of a normal storage and synchronization mechanism between devices - changing their devices, people just lost statistics. Having picked up a backend in the Azure cloud, we solved this problem by developing a uniform format for storing and synchronizing data about the activity of bracelet users. We plan to use the IOS 7-9 , Android 4.4.2 , Windows Phone 8.1-10 platforms , and it is important for us to concentrate on the development process, not on infrastructure issues, and the cloud in this sense is a great helper - you can raise the backend for storage without much effort. And data processing and, and, considering that we are in the program of support for start-ups of Microsoft and had communication with technology specialists, Azure, namely Azure Mobile Services, became the logical choice.
')
Despite the fact that data on mobile devices is stored differently due to different architectures, we have developed a single format for storing data in the cloud, and integrated into the SDK at the API level.
These are separate tables for storing various per minute characteristics of such activity, sleep, exercise, and human performance indicators - the number of calories burned, BMI, etc.

The third problem we identified is the ability to access statistics and infographics through the browser, which is a more convenient format for analysis. Also, access to anonymous statistics on our devices will allow you to quickly find errors and improve our product.

Solving these problems, as we see, will improve our competitiveness and we will be able to provide better service to customers. Which is a logical way should affect sales. The important point here is also that we are well aware that building such an infrastructure can cost us 100-200 tr. per month at full load and similar expenses may be relevant for us if the price of support for one user is no more than 10 rubles per month. But these are quite real indicators for us in a year.

Future plans

Our task is to implement 20,000 - 30,000 devices of the first 2 models. In the following models will be made bias on the sports, fitness and medical components. We will load special firmware into bracelets to fix certain movements, and also add the functionality of workout downloads from the application to the bracelet. The cloud along with the development of products will play an increasingly important role.

Smart bracelet on Boomstarter

Our hypothesis was that through the crowdfunding platform we can test the market without high costs. Now we are working on developing a bracelet on a new platform with a screen and heart rate monitors. We took the Nordic nrf51822 as a basis . On the other hand, we had certain difficulties with the sale of the current version of the T-Band bracelet due to the appearance on the Russian market of the cheapest Chinese bracelet Xiaomi Mi Band.



We have long been offered to make a bracelet on the basis of a capsule design patented by a Chinese factory, which is already a supplier for several bracelet manufacturers. However, we did not want to use an original design, and our Cortex platform was not very well suited for such a form factor. And here everything coincided - using the Nordic nrf51822 and this capsule, we could make an extremely affordable, high-quality and stably working bracelet. But in Russia, a significant part of the retail price is occupied by marketing costs, so we expect to get a similar product from the launch of the project at Boomstarter how much this product is interesting to the market and whether we can reduce the marketing costs, which now reach 1,200 rubles per customer.

This is our first experience in crowdfunding and we tried to prepare thoroughly. Studied close projects, prepared a marketing plan. Only the process of layout and revision of the text took a week of work of 2 specialists. At the same time, just a few days after launch, I already wanted to redo much.

However, Boomstarter is a very interesting experience. And to the extent that this platform was considered not as a place where you can get money for the launch, but as an advertising platform, then a new model will be in any case. This is our fundamental difference from most projects on Boomstarter. Why should you participate in this story through Boomstarter? Everything is quite simple, firstly, when we release our new bracelet to the wide market - it will in principle be impossible to buy it at this cost. Well, secondly, we will interact very closely with our first users on a new bracelet, and for each user this will be a real opportunity to bring something to our product. Help us make a bracelet that will suit everyone. Isn't that interesting? Is this not the story in which you want to participate?