MADI uses Ansys VRXperience to test unmanned vehicles at a virtual training ground

2021: Using Ansys VRXperience

MADI uses Ansys VRXPERIENCE to test unmanned vehicles at a virtual training ground. This was announced on August 18, 2021 by KADFEM Si-Ai-Es.

The Department of Organization and Traffic Safety of MADI was one of the first in Russia to engage in intelligent transport systems (ITS). Within the framework of the national project "Safe quality roads," MADI specialists developed the concept "Digital Road Model" (DMDD), which, together with cooperative and classic ITS, will ensure traffic safety and increase its efficiency, including for highly automated transport (VATS). To develop a prototype of a digital traffic model at the MADI Smart Road polygon and test site, MADI uses the capabilities of Ansys VRXPERIENCE Driving Simulator.

The digital traffic model will use the data collected in MADI, which will increase the efficiency of navigation functions on the VATS lidars and will allow you to train the algorithmneural network, including for recognizing markings, signs, etc. The DMDD can build a safe route taking into account current traffic conditions, including accounting for accidents, the state of the roadway and the effectiveness of the entire transport system. The creation of an unmanned transport system along with the development of road infrastructure will solve a number of problems - from safety issues and reducing the number of accidents, aspects of cybersecurity to improving the environmental situation and increasing throughput, making the lane more efficient by about 2-5 times (10 times on some sections of roads).

MADI specialists use Ansys VRXPERIENCE Driving Simulator in the digital model of the landfill to develop scenarios for the movement of WATS in normal conditions, abnormal situations, as well as in the background for information support of the driver during manual control of the WATS. The model includes the same parameters as the real road, including placards and technical means of organizing traffic (signs, markings). A virtual traffic light and a service were created to inform the car of the recommended speed. This allows you to increase the throughput of an adjustable intersection, avoiding delays at the start, saving time, as well as reducing the risk of accidents and congestion.

In the same model of the landfill, MADI specialists conduct a study. 7 scenarios were selected in which it is not possible to obtain the necessary information about critical facilities using VATS vision tools, which can lead to an accident. One (in a number of scenarios - two) unmanned vehicle due to visual obstacles does not see part of the road scene and must reconstruct it to avoid an emergency with other road users. To do this, the DMDD must collect information from other sources of dynamic obstacle detection and transmit it to the PBX. The experiment allows you to determine what accuracy the DMDD needs so that an unmanned vehicle can use this information to determine obstacle paths and plan its route to avoid an accident.

MADI specialists used Ansys VRXPERIENCE to set the basic scenario - creating an emergency. This is a rather difficult task, since an accident is a combination of many circumstances that had to be taken into account and create conditions under which both vehicles could not see each other before the emergency. The creation of these conditions reduced the number of physical experiments and saved about 20% of the time of specialists. Ansys VRXPERIENCE was also used for initial hypothesis tests. For the script experiment, 3 hypotheses were considered for each scenario and the best option was chosen from three.

We decided to use Ansys software solutions in this project, since they have several very important advantages. Ansys VRXPERIENCE is a comprehensive product that allows you to run most of the project in a single system and not switch to other solutions when new tasks arise. This, of course, allows you to significantly save time, finance and other resources, including for training, "comments Andrei Vorobyov, deputy for scientific work, head of the department" Organization and Traffic Safety "MADI, as well as deputy head of the MADI Competence Center for ITS.

The virtual landfill developed by MADI at Ansys VRXPERIENCE is also used to test the main functions of an unmanned vehicle before launching on the route. As of August 2021, MADI is working on two such projects. The first of them is made on the basis of a parking lot - a car located in the contour of the parking space control system of the city of Moscow. It automatically records the numbers of cars parked on a predetermined route, and determines violations related to parking rules. The project was carried out together with Mostransproekt and ROSDORNII.

The second pilot project is related to the pandemic and was carried out together with Mostransproekt and the Moscow City Innovation Agency on the territory of the First City Hospital named after N. I. Pirogov. An unmanned vehicle transports medical samples, including tests on COVID-19, from the place of their collection to laboratories. The territory of the hospital is quite extensive, between buildings and laboratories a large distance is 500-800 m. The appearance of an unmanned vehicle saved time, increased the efficiency of medical personnel and increased the level of discipline of internal processes due to the strict schedule of movement of the unmanned vehicle.

Unmanned vehicles are actively developing both in our country and around the world. They are becoming increasingly popular in various industries, such as aviation, railways, agriculture and many others. On the one hand, the creation of such vehicles is extremely promising and carries great economic potential. On the other hand, it has not yet been possible to achieve their complete safety due to the huge number of all kinds of scenarios on the roads. So, the sensors of an autonomous car are greatly influenced by visibility conditions, and weather conditions affect the dynamics. In this regard, developers recognize that conventional testing methods are not enough and the use of virtual testing is necessary. With the support of KADFEM Si-Ai-Es specialists, MADI employees actively use the Ansys VRXPERIENCE platform to test their autonomous vehicles. Virtual testing at Ansys allows you to evaluate the behavior of an autopilot taking into account the perception of cameras, radars and lidars obtained by accurately physically modeling scenarios, as well as maneuvering the car taking into account the work of systems and the quality of the road surface, "comments Alexander Kruglov, head of the Autonomous Transport and 5G networks, KADFEM Si-Ai-Es.