Siemens has announced SCAPTOR, the latest expansion of the Simcenter portfolio for autonomous vehicle (AV) performance engineering. The company claims it offers a complete, integrated solution for advanced driver assistance systems (ADAS) data collection.
What’s an ADAS?
ADAS is essential to developing AVs, which are undergoing intense development. While AVs are generating a lot of buzz, grave concerns about driver and passenger safety persist. ADAS development will go a long to alleviate the public’s hesitation about AVs and help make them a reality.
Adaptive cruise control is one example of an ADAS feature that addresses AV safety issues. It helps the AV maintain an appropriate velocity and safe distance from surrounding vehicles. Another example is the ADAS collision avoidance system, which detects traffic conditions that could lead to an accident and alerts the driver.
An ADAS can also analyze a driver’s breath and prevent a vehicle from starting if alcohol is detected, for instance. Night vision systems enable vehicles to detect obstacles like pedestrians or other moving objects more quickly in the dark or during poor visibility conditions. In addition, driver drowsiness detection systems can help prevent accidents caused by fatigued drivers. ADAS emergency driver assistance alongside lane departure and wrong-way driving warning systems can step in to control the vehicle if the driver becomes incapacitated, drives the vehicle off the road, or is driving in the wrong direction.
Along with helping autonomous and semiautonomous vehicles operate safely, ADAS can reduce the cost of road tests. Without an ADAS, AV road tests could take decades and cost tens of millions of dollars over billions of miles. ADAS computer simulation lets engineers test as many as hundreds of thousands of driving scenarios to optimize the hardware, sensors and the software algorithms.
Sensors on an autonomous test vehicle could include short- and long-range radars, low- or high-resolution cameras, or even LiDARs (3D laser scanning sensors). The enormous datasets these sensors generate must be analyzed in order to simulate millions of driving scenarios and so optimize AV warning and safety systems.
SCAPTOR
Siemens’ SCAPTOR, the new addition to the Simcenter suite in the Xcelerator portfolio, is critical to enabling the high-speed capture, storage and replay of in-vehicle ADAS datasets.
SCAPTOR enables time-synchronous visualization and replay of raw sensor data. Data from radars, LiDARs and high-resolution cameras is recorded, visualized and replayed in high-fidelity at ultrahigh speed. As storage requirements grow, SCAPTOR can scale to adapt. It is also ruggedized for continuous operation in extreme environments.
The Xcelerator portfolio integrates the entire Siemens Digital Industries Software portfolio with embedded tools, databases and solutions that span product design, development and manufacturing.
Now under the Xcelerator umbrella are these formerly separate solutions:
- MindSphere—Data-driven development of industrial IoT and smart manufacturing
- Siemens Cloud Solutions—Massive scalability and flexibility realized
- Mentor—Cost-effective electronic systems, components and products
- Mendix—Enterprise-grade apps and simple software tools to increase productivity
Simcenter software, which is also a part of the Xcelerator portfolio, focuses on system simulation that helps a client predict performance across all critical attributes during the entire product life cycle.
The capabilities of Xcelerator and Simcenter are already numerous and massive. However, additional hardware and software enhancements are needed for the development of ADAS.
Testing AVs is generating data measured in the terabytes. It is crucial to bridge the gap between vehicle development and big data management. The development of safe and reliable AVs requires realistic, high-speed, high-fidelity, raw data from multiple sensors. These data will be essential for training the algorithms and eventually validating and verifying the autonomous functions and the ADAS.
SCAPTOR provides additional capability which enables Xcelerator and Simcenter to execute a closed-loop development process that proceeds through data capture, analysis, exploration, verification, and validation before looping back to data capture. This solution encompasses the physical and virtual aspects of AV development, such as in-field data collection, 3D modeling, and virtual testing.