• +66 (0) 76 670 195
  • +66636502456
  • This email address is being protected from spambots. You need JavaScript enabled to view it.
  • Mon-Fri 8:00-17:30

Simulation Is Key To Real-World Autonomous Driving


One is a computer simulation that allows Ford to test its technology for autonomous driving in a safe and controlled environment, according to a new Ford report.

Using a computer simulation using data from Ford's vehicles, the virtual vehicles traverse a virtual environment created using computer simulations to study how they operate under real-world conditions such as traffic lights and traffic lights.

The data from these tests provide a snapshot of how Ford's self-driving vehicles behave before they are deployed in the real world. The simulation offers the possibility to study systemic performance in a way that would be difficult or even impossible for a human driver. It gives an insight into why it has become so important for automakers like Ford to work on developing and testing self-driving vehicle technology for autonomous driving.

The system trains driverless cars in simulation prior they hit the road

The control of an autonomous vehicle is largely based on roadways. Invented at MIT, the driverless car training simulation system creates a photo-realistic world with unlimited steering capabilities that help cars navigate a variety of worst-case scenarios as they cruise down a real road. Based on this data, they learn how to mimic safe steering control in a variety of situations.

It also makes it possible to rigorously test rare and dangerous scenarios that are difficult or impossible to hit in the real world. It is rare that you are forced off the road into the oncoming lane and almost crash, but it is also rare that you almost crash yourself.

Drive Constellation allows us to develop a physical car, but building an autonomous vehicle requires much more than just the physical vehicle itself. We ensure end-to-end testing, from subsystems to complete vehicle integration tests.

The challenges of developing autonomous vehicles during a pandemic

The division at the forefront of Uber's autonomous vehicle project will retain a team that will continuously expand testing in the Uber simulator, which is based on test tracks and road behavior data, he told VentureBeat. ATG engineers use a web-based interface developed by ATGs in collaboration with the Uber Data Visualization team to see how the car simulation perceives the virtual world. The software of the self-driving system runs in a virtual environment, to which continuous adjustments are made. 

DataViz, which provides data visualization, analysis, and data analysis services for real driving scenarios as well as data analysis and analysis for virtual reality.

Its Contribution to Evaluate Highly Automated Driving Functions

To achieve this, we first introduce a traffic scenario that serves as the basis for a real driving scenario with real driving scenarios. The description is based on the ground truth measured by Differential GPS, which is simulated by a test vehicle deployed in the real world, with the help of real drivers and real vehicles. This description, based on the environment subsequently created, leads to a reprocessed track that can be compared with the corresponding "real" data to illustrate the resulting behavioral changes.

In order to make the behavioral changes interpretable through the evaluation process, sensitive risk values are used to contain the recycled quality of the selected description as well as the real data.

Simulation Testing in Autonomous Driving Development

With the simulation, developers are able to test and repeat extreme situations to achieve deterministic operation and consistent system response. The innovative simulator for autonomous driving VR can automatically generate a base. To start simulation tests, the developer can build a virtual environment by mapping and importing real-world driving scenarios and adding characters, artifacts, trees, and traffic signs.

Tests are carried out in cooperation with leading innovators of autonomous technology such as Google, Tesla, BMW, Audi, Mercedes-Benz, Ford, Nissan, Toyota, and others.

Adams-VTD Integration:


The first steps towards the realization of driverless cars will take place publicly in the virtual world, but not under real conditions.

Companies are using highly detailed computer simulations to optimize their vehicles before putting them on the road, according to a recent ABI report. The software-based technology allows engineers to drive in a virtual world with up to 10,000 miles of driving experience and try out all sorts of scenarios they could ever go through in the real world. Waymo, the market leader in autonomous driving miles, has covered as many autonomous miles in its cars as in all of its vehicles, totaling 1.5 million miles.

NVIDIA's Drive Constellation System relies on the computing power of two different servers to create simulations of millions of kilometers of roads that cover a wide range of traffic conditions.

The first server runs software that simulates the sensors of self-driving vehicles, while the second processes the simulated data as if they were actually driving on the road. Cognata offers AI-based traffic agents based on drivers "behavior in the real world, enabling the creation of highly realistic road scenarios. Researchers at the Massachusetts Institute of Technology have developed a system that creates real-world traffic conditions for a wide range of road conditions.

Autonomous vehicle control systems rely largely on roadways, and controllers can learn from this data how to mimic safe steering and control in a variety of situations. Designed to help themselves - driving cars to learn to navigate through numerous worst-case scenarios as if they were rolling over real roads.

Please publish modules in offcanvas position.

getLinks(); ?>