Creating an environment model for automated driving
Detection of the vehicle environment using a combination of cameras, RADARS, LIDARS and cloud-based data can help automated trucks orientate themselves in their surroundings
Reliable perception of the vehicle environment and its precise evaluation is a basic requirement for highly developed driver assistance functions and automated driving. Technology company Continental is developing an environment model that captures the vehicle environment using various sensors. Data from sensors such as camera, RADAR and LIDAR is combined with additional information such as the profile of the route ahead. The data is evaluated and interpreted by an intelligent control unit and Continental’s assisted and automated driving control unit (ADCU), a high-performance computer. A complex and detailed environment model is then created from the resulting data. The ADCU creates an environmental model more than fifty times per second by linking information from the individual sensors and the various applications.
Continental argues that only through this view of the vehicle environment can vehicles such as trucks orientate themselves in their surroundings and make driving strategy decisions, for instance, by recognizing possible driving corridors. However, the information can be configured in many ways, and manufacturers have the flexibility to integrate the functionality into their systems.
Continental offers various radar sensors and cameras for environmental detection. Its development engineers are working on a 3D Flash LIDAR for passenger cars and commercial vehicles. The advantage of using a combination of different sensors is that it provides a more reliable and accurate view of the environment. Each sensor has its individual strengths and detects different environmental parameters. In addition to sensor data on other road users and static objects such as lane markings and traffic signs, more data also flows into the model by means of connectivity technologies for vehicleto-vehicle (V2V) and vehicle-to-infrastructure (V2X) communication; HD maps and GPS, for example, provide exact positioning data about the vehicle. Systems like the dynamic ehorizon and traffic data from third-party providers can also take the entire traffic situation into account, such as a traffic jam up ahead or a mobile construction site. This creates a reliable image of the vehicle environment.
Dr. Michael Ruf, head of Continental’s commercial vehicles and aftermarket business unit, said: “The environmental model complements Continental’s portfolio of components and subsystems for vehicle environment detection. With our sensors, our applications for vehicle connectivity and intelligent control units for automated driving, Continental will in future be in a position to offer its customers everything they need for the reliable detection of the vehicle environment, from a single source. To achieve this, we have used our experience in automated driving functions for passenger cars, more than ten years of experience in sensor applications for driver assistance functions for trucks and our in-depth systems expertise.”
Merging of information from a camera, LIDAR and RADAR to create an accurate view of the environment.
Perception of the vehicle environment by an automated truck.