Interview of TAdviser with the CEO of ISS company Oleg Krupenko

Oleg Krupenko: The idea of the smart home consisted in creation of additional comfort, simplification of control of the difficult home equipment. The smart home offered not so many opportunities - temperature maintenance, work from one button. So entered the building – light lit up, the TV was turned on. Also there were 3-4 modes. For example, Guests - complete lighting, Protection - simulation of finding of people in the house that nobody climbed.

Everything was very expensive and is always individual. The person at first built the house, selected to himself boilers, ventilation, and only then began to think of the smart home. Every time was necessary to solve absolutely unique problem. It was necessary to save expensive repair, and wireless technologies came to the rescue. And they bear for themselves new problems too: any sensor has the limited term of working capacity, it is necessary to change at least power sources. The smart home both was available to rather secure category of people, and remained.

Though there are rather cheap options. It offers from Intel, Samsung, Apple which already created an ecosystem of devices they have a managing controller - normally it is the computer or the smartphone. Also it is necessary only to install the application and to configure. In the future the control of the smart home will be undertaken by these companies. To them it is simpler as they develop infrastructure, do refrigerators, lighting. But it is necessary to understand also that all of them will support the proprietary standards. Apple will never work with open source software, at the same time its install base is huge, and in this direction it will strongly dominate. Equipment manufacturers will set standards of management and to define conditions.

Oleg Krupenko: In the long-term future Smart home will be as in fantastic movies when you have natural language of communication with the assistant, it executes voice commands and informs you by voice on what occurs. The user needs Smart home which communicates with you, but not just manages the refrigerator. He can ask you - what you want to change in a food allowance next week? And further a system itself should find the correct products in shop, look whether there is an empty seat in the refrigerator.

Without natural language of communication it is heavy to ordinary person to understand commands which should be issued. Even the simplest interfaces are control panels, buttons - now are quite difficult. The beautiful program interface will not save here, one more abstraction layer which considerably will simplify life is necessary. Development goes on this way – there are already TVs on voice actuation, the assistant to Siri.

Oleg Krupenko: We came from other market, were engaged in system integration. Our customers - the large companies with a large number of a geodistributed infrastructure, we solved a problem of management of it. Created monitoring systems, dispatcher stations, command centers.

When there was good telecom infrastructure, we came to a conclusion that it is possible to apply the practices not only in the form of the closed solutions in corporations. It is possible to create reliable service based on Internet technologies. The paradigm of Internet of Things assumes that you have a global cloud to which the objects capable to interact with each other are connected. We began developments 3-4 years ago.

At the standard smart home there are 2 I.Qs. The first - the device, in it some logic is sewn up. For example, in the same conditioner there are different operation modes, rotational speed, the movement of blinds. The second – logic at the level of the controller. It aggregates data from sensors and devices, in it algorithms of their processing are implemented. We offered some more I.Qs.

In our systems very powerful intelligence is taken out on the controller. It can work in a standalone mode, without connection with a cloud service and execute the local management program. But it also transfers data of measurements and all history of events in the normal mode upward. There in a cloud one more I.Q. – network management of controllers, working off of unusual situations with involvement of operators – the people making decisions, a possibility of manual control by an object is implemented.

We called such architecture Smart Office because this technology is well applicable for office solutions. As a rule, both retail, and banks, and state institutions - do the offices unified, they have a similar structure of the equipment, clear rules of work or the requirement of standards on comfortable Wednesday, maintenance of the light mode in the city, etc. Our technologies strongly differ from the smart home as there is a need of management of a large number of objects.

The companies with big network of offices are faced by two serious problems: reduction of operating expenses and creation of comfortable conditions for employees and visitors. Energy consumption optimization, minimization of manual work and human factor, constant control of infrastructure yield fast and tangible result.

Oleg Krupenko: It is possible to cut down considerably the staff of engineering service, or expenses on outsourcing of these services. The person who will go round objects is not required and to periodically check how everything works. In case of some incident the specialist coming on a call will precisely know a cause of defect. Now often primary diagnostics "by phone" strongly differs from what the engineer really faces on an object. Figuratively speaking called the electrician, and the fireman is necessary.

Our system allows to diagnose the reason of change of the correct operation mode with high confidence figure. If temperature sharply went up, we check how the systems of a thermal node, the exhaust ventilation, conditioning work. It is logic of the third level, there is a correlation of events from different sources. At our managers all information before eyes, they can precisely tell what needs to be done on an object. And moreover, if it is possible, the operator can correct a situation far off: reboot the equipment, turn off the light, close a door.

We began with engineering services, but gradually our system turns into an end-to-end system of management of an object. We integrate into it security systems, video surveillances. System value for the company sharply increases. Not only the number of engineers is optimized, but also the security service falls under the sequester. Often on an object function of the security guard - it is simple to click in time. He is not armed, does not show actual resistance, only fixes the fact of violation and clicks. We replace it.

Oleg Krupenko: No, our task not to save to the maximum, and to optimum use energy resources – to perform power management. Let's take the banking sector. If you have an insufficient illumination level, clients suffer with filling of forms, clicking of buttons, especially elderly people. If oxygen does not suffice, all work slowly, clients slowly communicate, operators quickly are tired. Cold indoors or too hot – everything affects working capacity.

It is important to support comfortable conditions, but to do it by the minimum quantity of resources. Here our technologies really help. In a pilot project for Sberbank we reached the economy confirmed with employees of the bank to 18% on the electric power. Now Sberbank continues to implement our system, characteristics are confirmed. Last year implemented 500 objects, this year will be even more.

Oleg Krupenko: The controller can execute transfer of an object to emergency operation if several critical sensors worked. For example, there is a flooding or the fire. Then it executes complex transaction: gives a task for blocking of electric inputs, blackout if there is a gas - that gas, blocks water, opens premises, etc.

We consider different cases - the terrorist threat exists, and there are scripts of preparation of an object for this action. Time for preparation of an object after detection of such danger is usually calculated seconds. People do not manage to react correctly. For example, in case of threat of explosion it is important to unblock competently doors, to open as it is possible more widely all windows - it reduces influence of a blast wave. The person will not be in time, automatic equipment should work here. It should notify people, to disconnect something, something to include - here the controller should fulfill such difficult scenarios locally.

The controller can also catch situations when fight of different subsystems begins. For example, the ventilation driving warm air, and the conditioner at the same time works. There is a double overexpenditure of energy. The controller compares as different devices what temperature outside and indoors, and understands because of what it is high work.

Oleg Krupenko: Problem of a system at this level - to issue to the manager all necessary data and signals that he saw and understood a situation, took those actions which a system should not execute without confirmation of the person.

There are transactions which are executed automatically, these are group politicians. For example, you control energy consumption of office and see that it begins to approach tariff restriction. Having passed through it, it is necessary to pay on a bigger rate. You can change an operation mode in advance - to pass into the economical mode. The comfort will decrease by couple of days, but money will be saved. Or, for example, operating time of offices changed - it is possible to recustomize algorithms one button everywhere.

And there are unusual situations. Example: the fire warning worked. If our system is not used, usually the signal goes to fire service, from there come firemen. They drive up to bank department, and there is cracking of the ATM, for example. The manager should understand our system that occurs, and he for this purpose has all tools. He can request video signal, look whether there is a noise, smoke, lighting.

Besides, our service has fourth I.Q., the fact that now call Big Data. We collect more than 100 parameters from an object and we store them throughout long time. Analyzing data, it is possible to find out many things. For example, aberrant behavior of an object. It is so possible to catch illegal connection to the power supply network easily.

Using Big Data it is possible to detect the inefficient engineering systems. See that energy consumption of the device big, and effect small. We see degradation of the same conditioning systems if they were incorrectly serviced. Such things on big intervals become visible.

Based on collected data it is possible to train a system to work. So, to opening of office in it there has to be already comfortable environment. It is necessary to include in advance all engineering systems that to opening there was correct temperature, humidity. For each office an interval when what to include, is unique. It is impossible it is model to calculate as it depends on many factors. But we accumulate Big Data of behavior of an object. A problem of an algorithm – to find similar conditions and to solve when to include these systems. Difficult algorithms, difficult mathematics are for this purpose necessary.

Oleg Krupenko: We buy only a small amount of component parts in Taiwan. And all equipment, sensors, sensors gather here in Russia. We initially wanted to make a domestic product therefore our circuit engineering. The payment our original, becomes at the Russian plant. The quality of our equipment is better, than the Chinese assembly, at the level of Taiwan, Korea. Experience of application shows that percent of failures very low. We make small batches, this year, for example, ordered about 3 thousand devices.

Our controller quite powerful. In it the ARM9 processor 433 MHz, 256 MB of RAM, for data storage are used a SD card of 4 GB. In terms of protocols at the physical layer we support RS485, Ethernet, WiFi, analog RS232. In terms of logic – ModBus, KNX and the protocol of management of lighting.

Temperature sensors, illumination, CO2 level, leakages, existence of current and also functional relays for device management integrate in network using physical cables. On business objects we do not risk to put the sensors managed through open air, we apply them only in the most extreme cases.

In the equipment we use open source software. Our task - to create the platform which will conform to standards of Internet of Things.

Oleg Krupenko: Now we are aimed at the companies with a large number of geographically spaced offices, first of all at financial institutions. High potential in retail. There a system will help not only to save the electric power, but also to save from the direct losses connected with non-compliance with temperature moisture conditions of storage. It belongs also to drugstores. All large network retailers suit our system.

One more business - gas stations. Now the refueling station is a difficult object where there is both a retail, and complex service. And object power-intensive. Well and city utility objects – hospitals, schools, state institutions.

Technologies can be used at cottage construction, a house system with the new standard on energy efficiency. At a design stage it is possible to put infrastructure of sensors at once, to provide the correct electrodistributing. And if the owner of a cottage wants, he will be able always to install the equipment, to be connected to service to monitor energy consumption and to manage the house.

Oleg Krupenko: The technology is reproduced, repeated. And here experience is unique, it cannot be taken from others system, all the same it is necessary most to pass through all errors, having stepped on many traps and having found the correct track. The fact that we accumulate this experience - huge competitive advantage.

We want to attract partners. In architecture we have a part - the platform and a part - modules. We will be engaged in development of these modules together with partners. Our company is open in terms of cooperation. We expect to create specialized solutions for verticals: municipal services, municipal economy, retail. To us it is important that experience of implementation turned into development of our software.