Anti-son Contactless solution for monitoring the condition of drivers

The main articles are:

• Computer vision (machine vision)
• Transport telematics (global market)
• Neural networks (neural networks)
• Big Data
• Biometric identification technologies

"Anti-son" is a comprehensive system of monitoring, driver assistance and analytics.

2021: Embed Next Mobile Operator SIM Cards

The Next Mobile operator with automatic flawless connection technology (ABS) and the KSOR company began cooperation in order to reduce accidents in passenger transport and improve the safety of citizens, which was announced by the latter on March 1, 2021.

Now Next Mobile sim-cards will be used in Anti-Son devices, which are already installed on the rolling stock of the Moscow Metropolitan State Unitary Enterprise and on all ground public passenger transport in Moscow, and are also involved in the national project "Safe and High-Quality Roads."

Next Mobile is a mobile operator in the Russian Federation that provides aggregation of all large cellular networks and automatically connects the subscriber to the highest quality connection. Next Mobile SIM cards with ABS technology allow the Anti-Son system to work smoothly, independently selecting the network of one of the three operators with the best reception at any point. We are confident that this cooperation will contribute to increasing the efficiency of the system and reducing road accidents and, most importantly, improving the safety of passenger traffic, - said Vasily Pisarev, Product Director of Next Mobile.

The system "Anti-son in real time using the camera analyzes the driver's condition, revealing signs of drowsiness and distraction. Having discovered the possibility of an emergency, the system notifies the driver with a sound signal, and also instantly reports this to the monitoring center. The Anti-Son system has already confirmed its effectiveness: the accident rate on ground passenger transport in Moscow in the first year of use - 2020 - decreased by 26%.

{{quote 'We are continuously working to make the system even better, more accurate, more reliable and more accessible. The requirements for external solutions and components are special - here reliability is triple important. During 2020, we conducted communication testing from Next Mobile in a variety of conditions, including underground transport and regions with low coverage, and based on the results we decided to use them on an ongoing basis, "said Denis Tsurko, Technical Director of KSOR LLC.}}

2020

Structure

As of September 2020, the complex consists of three development components of the Xor company:

• Observation device:
  • A video camera with an infrared filter is mounted in the driver's/driver's cab and monitors its condition 24/7.
  • Computing unit collects and transmits data (telemetry, geolocation)

• A neural network that learns from the collected and marked datacet (several million images and tens of hours of video) and is completed during operation on new material every day:
  • Analyzes the incoming video stream directly on the device.
  • After detecting signs of falling asleep or abnormal behavior, sends an audible warning signal to the driver.
  • The system notifies the dispatcher in the monitoring center.
  • Marks the desired area of ​ ​ the video for subsequent human analysis, including metadata (speed, location).
  • Saves data to replenish the data set and build predictive models.

• Monitoring Center (not only statistics "in the moment" are available, but also historical data are accumulated):
  • Rapid Response Dispatch Center.
  • Provides access to analytics and fleet information.
  • Risk profiling of the driver and development of recommendations for advanced training.

Types of classified events:

• Falling asleep/tired (eyes closed or open)
• Smoking/talking by to phone
• Driving a vehicle without a driver
• Interference in visibility/camera lens closed or turned off
• Long-term stops/deviations from the route

Development of the Anti-Son monitoring platform based on big data. Exit to the regions

In 2020, the next generation of the Anti-Son device was developed, on the basis of which a product line for commercial enterprises was formed.

The following tasks have also been solved:

• Development of big data monitoring platform (advanced analytics, risk profiling

drivers).

• Coverage of the Central Administrative District and access to the regions.

2019: 'Anti-son' integrated into transport video surveillance system

Holding "Roselectronics" State Corporation Rostec presented in June 2019 at the forum "Army-2019" contactless solution for monitoring the state of drivers "Anti-Son," integrated into the video surveillance system on transport. Development based on computer vision prevents falling asleep behind the wheel, and also collects and transmits data on the state and actions of drivers to a single monitoring center for the road carrier. The solution can also be applied on military equipment to control the condition of operators.

The DVR-like device is mounted in the vehicle cabin. When starting the engine, the equipment automatically identifies the driver's face. According to his facial expressions, eye movement and head turns, the device records that the driver falls asleep or is distracted from the road, and turns on sound and light signals. Information about the actions and condition of the driver is also transmitted to the monitoring center - a cloud service, where in the personal account the client can track in real time the movement of vehicles connected to the system and receive statistics on drivers.

"Driver drowsiness and fatigue are among the leading causes of serious road accidents. The new development is designed to reduce accidents during passenger transportation. The solution works correctly under any illumination and unstable communication channel, which allows it to be used for remote control of military equipment operators, "said Roselectronics.

The solution for monitoring the condition of Anti-Son drivers is integrated into the video control system for passenger transport developed by NPO Impulse JSC (part of Roselectronics). The algorithm used in the system for processing video data with an accuracy of 100% calculates the number of passengers from the image obtained from the cameras above the door aisles of the vehicle. The system integrates with the route network, analyzing the number of passengers at each of the stops, which makes it possible to predict traffic congestion. The solution is also able to identify the passenger's identity by removing two biometric parameters: quasi-3D facial image and iris. The NPO Impulse system has been used in passenger buses in St. Petersburg since 2017.

2018: Pilot projects based on the Anti-Son system

A number of pilot projects have been implemented together with several enterprises of the Moscow transport node.

2014-2017: Prototype and first tests

The Xor team of engineers undertook to solve the problem of monitoring the driver's condition using artificial intelligence video analysis. A prototype was created and pilot tests were carried out.