Smart home and industrial automation: EKF’s experience with Yandex Cloud

EKF is an international brand of electrical equipment, integrated energy-efficient solutions for power supply and automation of industrial enterprises, and civil and infrastructure facilities. Electrosolutions, which represents the EKF brand in Russia, has two production sites in the Vladimir region, a network of logistics centers, its own research and development department, and a testing laboratory.

Using Yandex Cloud infrastructure, EKF has created two new products: EKF Connect Home, designed for managing smart home systems, and EKF Connect Industry, for automating industrial facilities. Thanks to the ease of managing resources and of scaling and deploying cloud infrastructure, the company has accelerated the introduction of applications into production.

For remote control of smart home components, Electrosolutions, under the EKF brand, developed an ePro24 controller and a service based on the Tuya platform. In 2022, EKF decided to enter the market for smart devices with its products for property developers. To do this, they had to create a new application that would combine ePro24 series devices with virtual assistants.

To avoid infrastructure and regulatory risks, the company decided to move the entire project to one of the Russian cloud platforms. It was important for Electrosolutions to comply with the requirements of Federal Law No. 152-FZ. Therefore, the company sought out a Russian cloud service provider. In addition, localization would reduce the number of delays in the operation of smart home devices, which would increase system reliability. In addition to the security requirements for data processing and storage, the platform had to comply with OWASP project requirements, which helps manufacturers and developers identify security issues when creating IoT devices.

The company chose Yandex Cloud because the platform met all requirements and also provided the Yandex Managed Service for Kubernetes^® service for rapid deployment of containerized applications.

During the migration, they had to not only migrate the existing solution to the new cloud, but also to deploy infrastructure for two new products: EKF Connect Home for managing smart homes and EKF Connect Industry for automating industrial facilities. Both products have a microservice architecture.

EKF Connect Home allows users to control smart home devices via a mobile app or using voice commands. EKF Connect Home has entities (home, room, device) and rights levels: home owner and user with limited rights. User teams are converted into requests and are sent to the Apache Kafka^® message broker. They are then sent via flow, depending on the end services managing certain devices. User commands are displayed in the app. If automatic work scenarios are activated, the system performs actions on its own, for example, based on data from external services like Yandex Weather or Geocoder. When triggered, for example, when the alarm is turned on, notifications are sent to the user via Telegram or SMS.

EKF Connect Industry is a cloud-based analytical platform designed for dispatching and monitoring remote facilities. The platform supports two methods of transferring data from equipment, like programmable logic controllers and sensors. The first method is to transfer data via a Modbus gateway. It connects to equipment via Modbus RTU/TCP, polls it and transmits data via GSM, Wi-Fi, or Ethernet to an MQTT broker (EMQX) that acts as an intermediary between the object and the platform. Another method is to use an OPC UA server that connects equipment to a PC or PLC via Modbus RTU/TCP. The server collects data and transmits it via GSM, Wi-Fi, or Ethernet to the OPC UA client, the EKF platform’s microservice.

Both services are deployed in the same Managed Service for Kubernetes^® cluster, but are not connected to each other. All application data, including reports and statistics, is analyzed and stored in the Yandex Managed Service for ClickHouse^® cluster, which currently stores about 6 million records from various sensors. The storage period for such data is one year. The Apache Kafka message broker, deployed on Yandex Cloud, manages data reception and transportation, in particular, sending it to an online repository based on Redis^™. This data can then be used by customers to track parameters such as voltage. Data storage in Managed Service for ClickHouse^® allows for further analysis. The processed data is visualized using Python^™ 3.10 scripts and displayed as widgets on the application dashboards. EKF uses its own development for visualization. The platform has a built-in engine where users create scripts themselves to automate reactions and actions.

The company built a fault-tolerant application on Yandex Cloud for more than 1 million users.

After the launch of the project in Yandex Cloud, the MAU of EKF Connect Home quadrupled, and the number of active sessions in the app increased more than fivefold during one year of operation. Using the Grafana monitoring service, the Electrosolutions company monitors platform reliability using several metrics: the CPU and memory used in pods, the number of incoming requests per minute, active connections, and errors.

The company also created another solution based on EKF Connect Home and EKF Industry, EKF Building for the digitalization of residential buildings.