Digitising GAZ Group’s vehicle fleet. Vehicle fleet monitoring and management system

Client

1
manufacturing plants
0
all GAZ vehicles are connected to the GAZ Connect platform

GAZ Group is a Russian carmaker with a focus on the development and production of light commercial vehicles, medium-duty trucks, buses, power units and auto components. The company has 13 manufacturing plants in eight Russian regions, as well as sales and service businesses.

GAZ was the first Russian carmaker to embed telematics units in the serial equipment of its vehicles. From 2018, all GAZ vehicles have been connected to GAZ Connect, a single platform for accessing digital services via smartphone, tablet or desktop computer. GAZ Connect makes it possible for fleet owners to monitor the condition of their fleet and improve business efficiency by reducing vehicle operating costs. Indicators broadcast by the system include vehicle location, fuel level, coolant temperature, speed, fuel consumption, brake and coolant levels, oil pressure, on-board voltage and various other metrics. The service also includes other modules, such as communication with selected dealerships, discount programmes, roadside assistance and insurance.

Data exchange within GAZ Connect operates via on-board devices, which transmit data about vehicles in real time to a unified telematic data system. The system also assures the sending and guaranteed delivery of commands to on-board telematics devices. The delivery of these commands is carried out both online (when the vehicle is connected) and offline (via SMS) modes.

Challenge

Within the next two to three years, GAZ plans to connect 200,000 vehicles to the GAZ Connect platform, which will generate more than 40,000 messages per second. The monthly volume of telemetry supplied to the system of 200,000 vehicles is over 80 terabytes. The GAZ Connect platform needs to give users the ability to adjust all parameters flexibly and optimise them to gain extra economic benefits.

Reksoft’s specialists took on the ambitious task of modernising the architecture of the existing system to provide high fault tolerance and performance, ensuring the processing of 40,000 messages per second. An important requirement was the absence in the system of a single point of failure and load concentration, and ensuring the possibility of scaling the system above the specified load without modifying it.

Solution

Reksoft developed a high-performance system based on the most modern technologies to collect and process data from vehicles connected to the GAZ Connect platform, which includes a control subsystem for on-board telematics devices installed on GAZ vehicles.

Reksoft radically redesigned the architecture of the application segment of the system, eliminated dependencies between application services, redesigned the composition of application services, and added new system components to improve performance. In addition, Reksoft significantly optimised the infrastructure of the private cloud where the system operated through the efficient management of computing resources and ensuring fault tolerance. Reksoft supports Gaz Group in system operation and provides technical support. The system operates in a private cloud, which is based on PureApplication System, a virtualisation management system. It provides placement and reallocation of virtual resources on physical servers, dynamic allocation of computing resources depending on the load on virtual resources and, in the event of hardware failure, reallocation of virtual resources to available physical ones. PureApplication System also provides the ability to monitor the health of and load on physical and virtual resources.

The application part of the system is based on microservice architecture. Application services operate in the RedHat OpenShift cluster, which additionally includes Service Mesh components such as Istio and Kiali.

The OpenShift cluster provides automatic scaling of application services depending on changes in their loads, with effective load balancing between instances of application services, and automatic recovery of their performance in case of failure. Kiali visualises a diagram of the interaction between services with timing for all requests, providing the ability to monitor the status and parameters of the service network, and proactively identifying potential risks of degraded performance or service failure.

The high-performance MQTT broker IBM MessageSight is responsible for the primary reception of data from on-board telematics devices, as well as sending commands to devices that are online. It authenticates the on-board device in LDAP when it is connected, and receives and buffers data from on-board devices. Messages received from the MQTT broker are enriched with additional information, processed by application services and stored in a telematic database. In the course of processing, messages can be sent to service providers in accordance with their subscription to data from specified vehicles.

During the system’s modernisation process, special attention was paid to the issues of monitoring and preventive identification of risks associated with a decrease in system performance and component failures. To solve this problem, monitoring, logging and auditing subsystems were introduced.

The monitoring subsystem collects more than a hundred indicators that show the state of all elements of the system. The monitoring subsystem is based on the open-source product Prometheus, which provides centralised monitoring of equipment, virtualisation level, containerisation system, monitoring of network interfaces, traffic, database status and application services. Thresholds are set for each metric, allowing Prometheus to notify administrators promptly of potential risks and anomalies in the system. The monitoring subsystem reflects in real time the current state of each component on dashboards, and makes it possible to configure flexibly all monitoring threshold values through its interface.

The logging and audit subsystem provides collection and storage of a log of operations performed by the system and user actions. The subsystem provides tools for visualisation and analysis of the operation log, and also allows in-depth analysis of the state of system services and security events in order to further optimise it or resolve issues. A separate task required in the operation of big data processing systems is the development of data management processes. When developing data management processes, the DAMA-DMBoK reference model was used as a basis.

The main areas being worked out are:

  • data governance in terms of the organisational and role model, policies, data management regulations and regulations;
  • data architecture;
  • storage, archiving and deletion of obsolete data;
  • data quality assurance.
System advantages
  • high resiliency;
  • high performance;
  • automatic scaling of services depending on the load.

The system does not need to be stopped to allow routine maintenance, the replacement or addition of extra computing facilities or the updating of system software.

The combination of all these indicators allows the client to cope effectively with a large amount of data and solve other complex issues during the operation of the system developed by Reksoft.

System implementation benefits

The system developed by Reksoft makes it possible to increase the efficiency of resource use and reduce operating costs.

In 18 months of use, no critical errors and failures in the system have been identified. The system downtime over the operating period was less than one hour.

Today, Reksoft continues to provide ongoing expert technical support and assistance in solving all technical issues that arise during system operation.

Technologies

  • PureApplication System
  • RedHat OpenShift
  • IBM MessageSight
  • IBM Cloudant
  • PostgreSQL
  • Kafka
  • Redis
  • Prometheus
  • EFK
  • Java/Spring

Services

  • R&D
  • Testing
  • Systems integration
  • Implementation
  • Application support
  • Application maintenance
Menu