Shorten the perception development cycle of self-driving cars
Highly automated driving (HAD) development is proving to be much more complex than initial assumptions. Today's perception systems are often still not up to the task for SAE Level 4 and require further research and development. Successfully developing HAD perception requires mimicking human perception— recognizing objects in most lighting and environmental conditions and roughly evaluating their distance. Without an accelerated and thorough test framework, it would be an impossibly complex and time-consuming task.
Register now for this webinar to learn how Simcenter Prescan360 can speed up inventorying and diagnosing perception system issues by combining test cases, smart sampling, leveraging cloud parallelization, and new ways to digest many results.
Achieving a high number of simulations will not be enough if simulation engines cannot produce truthful results representing reality that can replace real-world testing. Simcenter Prescan360's offering of Physics-Based sensor models and libraries of material optics properties are the result of years of research, large laboratory measurement campaigns, and simulation performance improvements.
Now available for cloud or cluster parallelized simulation, these models open the way for perception or motion-planning and tracking machine learning. They also enable perception processing assessment for automotive suppliers and high-fidelity full-system verification for autonomous vehicle OEMs.
HAD verification based on human-directed testing is reaching its limits in front of an infinite number of scenarios to test. Simcenter Prescan360 enables new automated methods to direct scenario testing faster than iterative "simulate and review" loops without missing any corner of the scenario space.
Watch the webinar to learn how to bring high-fidelity sensing and large-scale testing coverage together to accelerate perception development.
ADAS/AD Product Line Manager