How the telematic service Smartdriving.io was created - 100% Russian technological startup

Smartdriving.io is a platform for collecting and analyzing vehicle telematics data. Telematic services is a very promising niche, which is now developing all over the world.

The company develops a product (hardware and software), which is installed in the car and collects data on its movement and technical condition.
')
The developers managed to create a complete solution with a modular architecture, which significantly expands the range of application of the product. Engineers develop and assemble the device in Russia. This means that for the first time in the Russian market of telematic services, a completely domestic development has appeared.

We talked to the founders of the company, who told us how they were able to launch their development into the market in record time. We discussed all aspects of the development of the project - from the development of software and hardware to finding customers and monetization.

Why did you start doing this? How did you become a co-founder of the company?

Nikita Kasyanenko:

I have a shorter story. Michael and I have long been acquainted. Once we started working on a project that today is our competitor - this is the company Drive Journal. Misha served as director of development there, and I, as a third-party contractor, worked as a marketer, developed a corporate brand, style, marketing strategy. I also conducted research, studied the audience.

And in the process of working on the project, we came to the conclusion that this is a very promising niche, which has great potential, especially for our market with all its problems and underdevelopment in terms of automobile insurance. Therefore, we decided to make our own project.

When did you start the company?

Mikhail Anokhin:

In 2015. We gathered the team in June.

I started working in this market earlier. He had a certain background. Initially, in Russia, at the end of the 90s, companies began to appear that were engaged in monitoring transport and satellite security. These were technologies remotely similar to telematics. From there everything went.

At the beginning of my career, I worked for a company providing satellite monitoring services. Even then, everyone wanted navigation, GPS. At that time there were such huge Benefon-tubes in which SIM cards were inserted, and they sent SMS about the location.

When it finally became possible to track GPS movement in real time, monitoring companies and companies working in the field of satellite security were the first to use it. I worked in this area for some time, then I went to insurance - I worked there for 3-4 years. Then he returned to the security market in the company Avtokolator.

We did not even think about insurance telematics, satellite security was very expensive. The set cost 20-70 thousand plus 12-25 thousand monthly fee. When I came to Arkan to work, we first thought about insurance telematics there and did a project based on Israeli iron. Then the company 3S-Telematica appeared.

Soon came the first order: it was necessary to make a telematic solution for the company Ingosstrakh. This was done by the same technical director, who now works in our Laboratory. In itself, the decision was not bad, but expensive - 12 thousand cost, like a refrigerator. Then came the new leadership, a change began, and I left for R-Telematica. And there, Nikita and I just started doing Drive Journal. The project turned out to be interesting, but it had one important drawback: they also worked on foreign hardware. So, it was my third project, implemented on a foreign gland. During this time, I realized that foreign suppliers of iron often "exaggerate" its cost. And sometimes the functionality of their products is lame, although companies are very serious: even on the Nasdaq they bargain.

Therefore, we realized that the market is promising, but projects on it are implemented incorrectly. We saw a huge potential in uniting separate services (security, telematic units) into one device, making it cheap for Russian customers and selling.

The biggest question was where to find the developer. At that time in Russia, no one did high-quality iron. To those who tried to do this, there were questions, again, in terms of quality. Therefore, we were always in search of iron.

We had a talented but underestimated developer in mind. We have such a problem in Russia: engineers sell themselves badly. They have no such talent, but this is normal. He worked for more than ten years in large companies, so his experience is serious.

I suggest an analogy related to fitness trackers. There were solutions for professional athletes who watched their performance in training and so on. Progress has led to the fact that now any person on his hand may have a device that removes all indicators. Do I understand correctly that you are trying to make a certain fitness tracker for cars?

Nikita:

If you really simplify, then yes.

Michael:

But earlier those expensive devices did not carry the amount of functionality that our cheap devices now have.

If we talk about professional athletes, their fitness trackers are more expensive, high-quality, accurate. Our devices, despite the affordable price, are more functional. We have created a new piece of hardware from scratch: on new components, on a new base.

Why am I led to this topic? Because no insurance companies work with fitness trackers.

Nikita:

Our task is not to make people admire technology. We create for clients a service platform, the necessary part of which is such a device. Regarding fitness trackers and insurance ... A friend from America recently came to me. He said that he received a discount on Health Insurance because it uses Apple Watch. He works as an accountant in a medical company. Before you get an insurance policy, you need to fill out a questionnaire that indicates the use of fitness trackers, wearable electronics. So, they somehow use this data.

Who is the end user? Car owner?

Nikita:

Yes, but between us and the owner of the car there are still a few links in the chain.

How do you work with them? Why should the insurance company take into account your device and connect you to cooperation?

Michael:

The situation is this: in Russia, CASCO penetration into the OSAGO market is about 8-10%. In Europe, almost all insure - more than 80%. In Brazil, penetration is about 30%. So, we have very few people insured. This is all that insurers have achieved in 25 years. We have a lot of questions to these companies about the pricing, how they make products. In Russia, this institution is very conservative. Therefore, our product gives them the opportunity to recruit low-risk customers.

People who do not hit (or rarely hit) cars are not insured. Some dealers have a very clear correlation - there are as many insurances as cars taken on credit. If they sell 30% of cars on credit, they insure 30% of cars. The rest just go.

They need to build portfolios, because the CASCO penetration in Russia will only grow. In Russia, while the car is on the road, it carries great risks. Often people buy cars from Germany with very high mileage and rejoice: the car is in good condition. Because the structure of the roadway and the infrastructure as a whole there are better.

In principle, our accuracy is higher than in Europe. In addition, the risks on the road are higher. So, many insurers agree with us that it is logical to sell devices with limited mileage: one year it winds 40 thousand kilometers, the other 7 thousand. If a woman sits at home with a child and only goes to the store, it is generally 5 thousand kilometers. Why then should she pay for the policy as much as the taxi driver? Therefore, they do not pay. Most insurance companies are now working at a loss.

That is, people who drive rarely and safely do not buy full insurance because they cannot be offered an adequate price?

Michael:

When you get into an accident every five years, and you are offered to insure a car for 10%, you realize that in five years it will be 50%. Therefore, it is easier and cheaper to fix it yourself, if necessary. Here Nikita, for example, was not insured.
I remember there was a story when prices were raised sharply for everyone.

Nikita:

Yes, insurance companies compensate their losses by constantly increasing tariffs. And it turns out. What happens is such a "natural selection": either fraudsters are insured, or those who really badly drive and know that an accident will happen to them.

And what about dealers? They are interested in this solution in terms of installations in cars for example?

Michael:

Yes. Dealers have such an important task as returning the car to the technical center for service. If the client is not insured against the casco, he will not come there to repair the car. He will go to the next center and do everything for three pennies. If the client has insured casco, then the probability that he will return to the technical center to the dealer is much higher.

When the crisis began, dealers sold cheap policies for 1 or more insured events. However, they remained unhappy with these products: clients still did not return to them. And our devices, in addition to improving CASCO penetration, sell more policies, which are returned to them for repair, can also diagnose a car remotely and can invite a person to be inspected. The device collects data on the mileage, and when the time is right, it records the car for inspection, calls the customer and informs you that you need to come. In addition, it assesses the criticality of the malfunction: whether it is necessary to immediately go to the service or you can not rush.

And if we are talking about the regions? You live, for example, in Ufa, and bought a car for fifteen hundred kilometers from it. To go to the dealer you need to wind three thousand mileage. So, if the device showed some kind of error, you will think ten times whether it is necessary to go so far.

And now there is such a situation that car manufacturers close diagnostic equipment from others. For example, it is very difficult now to diagnose Lada West with unofficial equipment.

Let's now talk a little about the device itself. When my colleagues and I discussed this topic yesterday, there was a parallel with security systems for cars. We look at the technical description on your site, and there is a comparison with expensive alarms.

Michael:

First, we have extensive functionality. Most of what car alarms provide, we also provide. Secondly, the signaling systems that are currently on the market (Starline, MEGA-F, Pandora) are “two-way” or satellites (Caesar Satellite, Autolocator). They do not provide any insurance online services.



And how is the security issue solved in this case? You all send to some servers. I can not say that this is personal data of users, but in any case, quite valuable information.

We use secure transmission protocol. We comply with the law on personal data. We have leased servers in Russia.

Nikita:

In addition, we have a division: driving data is not transmitted along with the drivers' personal data.

How is the issue of iron security? After all, if an immobilizer is installed in the car, there must be a second key associated with the device that is connected to the diagnostic port.

Michael:

It all depends on the task. If you need to do autostart, then you need a second key. But not always: for some machines, immobilazera devices are sold, some have a corresponding set of commands - though this is rare for now.

Does this have any effect on the warranty?

Not. We do not do anything with the machine so that it can spoil it. Although if something breaks in the car, dealers always put the blame on the alarm. It always has been and will be. In fact, we can not impute anything, we do not spoil cars.

Can I fool this thing somehow? You have envisaged ways how attackers can neutralize unprofitable driving data in the process of reading or transmitting it?

Nikita:

We focused on linking it to the car as much as possible. The device itself constantly reads the VIN-number of the car.

What are the possibilities of cheating? Deception of the insurance company: the client can rearrange this device in another car, or in some kind of module. But we can scan data about the car, and also fix the process of disconnecting the device from the connector. As soon as this happens, a corresponding message appears in the system. We all control it and promptly report it.

Michael:

We try to weed out the scam right away. Now we are working on additional algorithms for working with the coordinates of the car.

It is good that you are trying to build such a process, when not the driver initiates the contact, namely you.

Nikita:

Yes, we provide the service. We do not develop technology for the sake of technology: we want to provide people with the highest quality service. For this we have a call-center in Petersburg. There round the clock technical support.

You exist for a year and a half. This is not the easiest development. This is not a CRM to sell. How are things at the moment?

Nikita:

The market is now segmented. What is our key difference? We are developing a modular system: we provide different types of services on the basis of one device. While in the market, such services are usually represented by different companies that do not compete with each other. There are satellite alarms and two way. In addition, there are companies that produce monitoring systems, work with b2b-clients and are engaged in car fleet maintenance.

And insurance telematics in the Russian market is just emerging. We are working in this segment, but in parallel we are developing in all directions. This is our uniqueness. In general, there are very few such companies on the market.

We are developing a retail business, working with car dealerships, selling car security systems, integrating existing insurance practices and so on.

That is, the sale of iron is not the only source of income for the company?

Nikita:

Not. We also make money by creating a subscriber base. From the point of view of marketing and promotion, our task is to increase the subscriber base. A person can insure and get our device in the insurance company. Or he will buy a security system and install it.

If we talk about the specific price, how much will the device and subscription for individuals cost?

Michael:

Often pricing is dictated to us by the dealer. Therefore, we in this will be very flexible.



Nikita:

15 thousand rubles now is the basic device. It allows driving analytics and some other features. To use a wider range of features, you will need to buy additional modules. We have a modular system.

Michael: For example, to add security features, you need to additionally buy a tag and a lock.

Nikita:

A label is an identification radio device that lies in the driver’s pocket and constantly “communicates” with the car. Each contact generates a new encryption key.

I am also interested in simple things - investment, revenue, team growth.

Michael:

Unfortunately, we cannot already speak about investments. This is secret information. About the growth of the team ... Now we have about 50 people already. This is quite a lot in a year and a half. Our competitors have smaller people. This is primarily due to the growth of our IT department. Now we are planning to increase the sales staff.

We do not have the proceeds yet. We just started to sell. We made the device for a year and a half - this is a very short time for such a complex product. When we meet with our potential partners, they do not really believe in it. We in such cases always offer to test it: install and ride.

What are the requirements for technicians looking for a job in a company?

Logical thinking, desire to develop, mat. apparatus, knowledge and experience in the profile direction of work.

What are your plans for 2017?

Michael:

First of all, we need to become self-sufficient by the end of 2017. We are now actively recruiting partners among insurance companies. By the end of the year, we should be represented in the largest possible number of such companies. So far there is such a situation on the market: each operator has a certain number of partner insurers.And we want to work with all insurers. That is, we hope that we will change the situation.

Of course, we plan to expand dealer networks in the regions. In addition, we will consider the possibility of entering foreign markets. We see that similar projects are globally competitive. There is, for example, the company Remoto. In Europe, there is the company Octo Telematics, which Vekselberg last year bought for 400 million euros. They are also developing all over the world.

While we have no competition with them in Russia will not.

How do insurers treat you? They do not see in you competitors?

Michael:

Fortunately, good. When we put together a team, we invited very experienced insurance specialists. I did some insurance myself. We know what problems they have there. Therefore, they see us as colleagues. They do not need to tell us what and how: we do everything turnkey, knowing the specifics of this area. We come to them and say: "Everything is ready, please: take and sell."

You explained that you started doing this because of the price and poor characteristics of similar products ...

Nikita:

It was a market need. We have seen that the Russian market is very specific. Why did we add security to this device? Therefore, this segment has existed here for decades. In addition, insurance companies have an interest in it.

Do you want to rise on the wave of import substitution?

Michael:

When we first started, this wave was just at its peak. Now I almost slept. I have a positive attitude towards import substitution. True, there is a problem with its implementation. Despite the fact that we assemble and design the device in Russia, we are forced to order components abroad. We use the Russian GLONASS, but even that, it is not going to turn out of Russian components. There is a big task of creating domestic microelectronics, but everyone is skeptical about this. We have no people, this is a complex science, therefore from scratch it is incredibly difficult to create. Not to mention the space investments that are needed for this.

But still, here we collect and develop our devices. In Tula there is a small, but modern, factory. We are happy with it. When we first started, we were told that in Russia no one could be trusted with such production: to do it either in China, or completely its own. It turned out that this is not true.

How and in what direction do you plan to upgrade your device? Will it get any extra features? Or is the maximum task now - to bring the product to some stable version?

Michael:

We had a bad and good scenario. Thank God, everything went according to a good script, and we managed to gain time. First of all, you need to hone the product to the ideal. We still have a marriage percentage. Usually, they always want to modify everything in this topic: add an LTE-modem, Wi-Fi, Bluetooth, a microphone ... let's insert a loudspeaker and scales.

Nikita:

I consider it a great advantage that we have created a system that allows us to increase functionality without interfering with the circuitry. We are flashing the device "over the air", and new features appear at customers overnight. There is no such thing that in order to introduce a new function, the client had to go to our office, change the device, and so on. From the point of view of iron, it takes into account all sorts of changes for the next few years.

Michael:

You can upgrade it, but it is not always necessary. Just replacing a modem from GPRS to 3G will already increase the price of the device by $ 20. But a very high data transfer rate is needed here.

Nikita:

When developing the firmware, we tried to reduce the amount of transmitted data as much as possible, so the current speed is enough for us.

What major players, besides Sberbank, are interested in your project?

Mikhail:

Now we are closing the contract with Rosgosstrakh.

We place high hopes on our auto industry, because competitors want to sell mainly on foreign cars. We believe that AvtoVAZ has prospects. For example, Lada Vesta is already selling well. This is a good car. Today we are focusing on this.

Sales of Vesta in million-plus cities will grow even more. There the product comes in very nicely because the inhabitants of these cities do not drive so much. Many young people are buying it now. While VAZ sells a little, but growth will be.

But in general, we work widely with all networks, not only with the automotive industry.

The need to create firmware separately for each car is dictated by the lack of a single standard for integrating different cars?

Michael:

Yes.Moreover, if you look at the Hyundai base and top configurations, then the command set will already be different.

Nikita:

We have to do this backing non-stop, we are constantly updating the corresponding database.

So you are not trying to get this data from automakers?

Nikita:

Not.It’s just impossible if you don’t have a formal contract with automakers, if you don’t sell something to him or some of his products.

Mikhail:

Moreover, even the existence of such a contract does not guarantee that you will receive the necessary data.

Nikita:

Because they have certain security protocols, they cannot disclose this or that information.



Would it be quite promising to sell your modules in grocery stores? How difficult is it to sell it as a box: he put it himself, stuck it himself?

Michael: It

depends. If you sell a security system - it will not work. If we talk about telematics - yes, there is a prospect in this: it is stuck in the standard diagnostic connector. Anyone can stick.

In addition to the market for new cars, insurers need a secondary market. There are 20 million customers in this market. Therefore, if you make a cool product, you can grow very strongly in a year. But while it is difficult.

There was such a telematics project “Important Insurance”. During the year they scored 20-30 thousand customers and entered the TOP-20 insurers. There, of course, there were questions about tariffs, unprofitability. But when the insurer only enters the market, it is usually always unprofitable.

Here I, for example, drove to the gas station, stopped. After that, the system tells me that a new trip has begun. But in fact, it is not: just the old trip continues after a pause. And it upsets.

Michael:

That is why we have analyzed a lot before launching the project. We initially thought that all assessments will be clear. If we make a mistake somewhere, for example, a sign is somehow incorrect, it can be challenged. This is what we planned for the future.

What difficulties were overcome during the development and production of the telematics module?

1) No one in the world has previously placed security, logistic, telematic functionality into such dimensions (OBD - II).

2) The accelerometer auto-calibration algorithm is a complex mathematical algorithm that requires consideration and analysis of a huge number of factors.

3) Reconstruction of the picture of an accident: with the help of our device, you can reconstruct the accident process, for this, the data is written to the ring buffer in a non-volatile, unmodifiable memory.

4) Interaction with all existing protocols used in the car CAN bus. Constant reverse engineering and the creation of new firmware for the device.

What stack of technologies is used on the side of software development and maintenance of a software complex? Programming languages ​​/ databases. What tools are you developing?

1) Microsoft .NET C #, Cross platform C ++, iOS Swift, Android Java;
2) Microsoft SQL Server, Azure;
3) Visual Studio, XCode, Android Studio, QtCreator.

Why did you start making your cartography if there are already a bunch of open solutions like Open Street Map?

Michael:

We have specific requirements for telematics cards. We need to account for hazardous areas. A map is essentially a database. There is a geocoder that works between the database and your queries. The question is simple: we need speed on every street. This will not give us neither Google nor Yandex and so on. Moreover, the more customers will use Google maps, for example, the more royalties we will have to pay this company. And the cost of the product is thus seriously increasing.

The card itself has its own database, into which we can add any objects - refueling, for example. If you get the status of "Fan" at the BP, then the system will help prove that you use only one filling. This is a very promising story.

Is there a long-term history of more autonomous driving? In particular, I mean the system "Plato". In fact, they are trying to do the same, only simpler: a scanner, a label, you are going to scan.

Michael:

The systems for heavy vehicles usually have very simple functionality. They have simple tasks: you need to see where it is, fuel level, 1C integration, metering, entry-exit and that’s it.

This is not our story. We can do all this. We have our own market, where we know what people can offer. Yes, we offer monitoring services for commercial vehicles, but you can install 10-20 cars for yourself.

But we are not going to get involved with Plato. There is a slightly different device, there is a standard in which you don’t need to dream anything.

And what do you say about the drone?

Michael:

We are watching this story, studying it. We don't have much enthusiasm here yet. This story is now quite bloated, it is easy to pull up money and investments there. But when she earns is unknown, especially in our country. Our road infrastructure is completely unprepared for such experiments. The markup for UAVs costs a lot of money. Any motorcyclist will tell you that there is such a markup in Russia, that ...

What is your dream about this company? What place would you like to take in the Russian market?

Mikhail:

We want to take a leading position in this market. This is not a dream, this is a plan.

Nikita:

What is important, we didn’t just make another decision, which ranked one on another, we are creating a new market. If we succeed, the auto insurance system itself will change. We want to make auto insurance more affordable. For the hull, for people, it means that you are either forced to make a bank, or you don’t need it at all.

Is there a goal to make it more understandable? Many people say that insurance indexes, models are opaque, it is not clear that at what point it is charged.

Michael:

We have a slightly different approach to this. In our application there is an assessment of the aggressive behavior of the driver on the road. This is not an insurance scoring, it is an assessment of behavior, where we give him points.

This does not mean that he will be denied insurance. The second part is already scoring. And our application can just explain to the driver his mistakes: where he exceeded the speed and so on. We do this specifically for the user.