UST: News

Transportation systems that are widespread today cannot do without human participation in management. Own "brains" of cars, airplanes and even trains are not yet sufficiently developed to trust them by one hundred percent. Nevertheless, automatic control systems are one of the main trends in the transport sector, which was confirmed by the exhibition InnoTrans held recently. Many digital solutions for traditional transport were presented there.

SkyWay project did not stay on the sidelines from the trend towards automation either: initially, the possibility to control traffic based on the readings from sensors using artificial intelligence was designed in SkyWay transport systems. There is a significant feature in SkyWay's intelligent control system that distinguishes it from competitors: it combines a large intellectual complex that will not only control the movement of transport modules, but also collect the required data on the technical condition of units, interact with the passenger and perform a number of other functions that are most often assigned to people in traditional transport.

Like a person, a vehicle needs information about the situation around it to take decisions. In SkyWay transport modules, the collection of such data will be handled by special sensors and detectors of several types. Information from them is collected continuously and sent to the control center, where routing tasks are compiled and adjusted. Moreover, the transport module is able to take particular decisions independently; only the most general instructions come from the central control point — specifying where and at what time it should be.

To control the traffic, the rolling stock is equipped with a combination of cameras and radars, which allow to determine more than 90 different types of objects, as well as the parameters of their movement — speed and distance to an object. The combination of "radar + camera" was chosen by SkyWay engineers intentionally — this set allows to recognize obstacles under challenging weather conditions and at the same time does not require such large expenses as the lidar (Light Identification Detection and Ranging).

The principle of operation of this complex is easiest to explain with a specific example: for example, a drone is approaching a moving unibus. Its image, as well as data on speed and position in space are transmitted to the on-board computer. According to the data from the camera, the machine vision algorithm determines that this is exactly a drone, and the data from the radar allow to calculate the speed with which the unibus needs to move to avoid a collision. The control system instantly makes conclusions and transmits data "to the wheels" — it forms and sends commands for specific machine units. After the on-board system has solved the problem, the data on the speed and position of the module are sent to the control center and the adjusted routing task comes back.

Changes in the route that are needed for traffic safety and accurate follow-up of the schedule can also be made on the basis of readings from sensors fixed on the infrastructure. The data from the cameras and sensors go directly to the control center, and from there, already converted into routing task adjustments, to all modules that move along difficult track parts. Special attention is paid to potentially dangerous areas, as well as to stations — additional cameras and more accurate sensors will be located there.

To determine the position of the rolling stock on the track structure, several methods are used at once — radio-frequency identification tags, GPS and magnetic sensors. The latter are used if increased positioning accuracy is needed — for example, when stopping at a station.

By the way, the technologies described above have already passed "combat debut" at EcoFest — manual adjustments to the computer were not needed for the entire day of operation in the automatic mode.

Sensors in SkyWay rolling stock are responsible not only for its movement — they also perform work that falls on the shoulders of depot employees in traditional transport. Various sensors of temperature, rpm, wear and other parameters collect information on the technical state of modules, send it to the control center, where the computer, referring to the norms specified by the manufacturer, decides whether it is necessary to send vehicles for maintenance or to allow their further operation.

A lot of diverse sensors and detectors are installed both in vehicles and at infrastructure facilities — for example, cabin temperature sensors, battery performance indicators, sensors on door opening, turn-outs positions, etc. The data that comes from them is processed directly by the on-board computer and are used to take specific control decisions: for example, when the battery overheats, the vehicle decides itself whether it needs to stop immediately, just slow down or it can continue running to the next station.

Data in a revised form get from the vehicle to the control center, where a special system collects the necessary statistics and decides on its basis what maintenance is needed for a specific module. It compiles a maintenance worksheet and lays down maintenance in the schedule.

All sensors, vehicle units and computers are in constant cyclic communication — changes in sensor readings affect the transport operation and the changes in the transport operation are reflected in sensor readings in their turn. All this is recorded by the computer, analyzed and affects the adjustment of routing tasks and maintenance plan. Due to this, the entire system is under constant automated control.