Autonomous Driving Simulator Improves Safety of Self-Driving Cars
April 30, 2019Researchers recently developed a new technology using photos, videos, real-world trajectory, and behavioral data that will provide a safer, more authentic self-driving simulation system. A computer scientist from the University of Maryland collaborated with a team of researchers to develop the Augmented Autonomous Driving Simulation (AADS), which allows them to test automatic driving technology and ensure that it is safe and reliable before allowing the technology to be tested in real cars on roads and highways across the country.
In addition to using photos and videos, the AADS system incorporates three-dimensional renderings that offer real-world data about pedestrians, bicycles, and other vehicles. This valuable data can be used to predict the behavior of other motorists and the positions of other vehicles and pedestrians. According to researchers, it is not easy to capture the way humans drive just by using mathematical models and the laws of physics. The research team reviewed the video footage they had and used the data to model driving behaviors using social science methodologies. This approach provided a more realistic and beneficial traffic simulator.
Challenge of Using Real Video Imagery
Researchers discovered that using real video imagery and lidar data for their simulations presented significant challenges. For example, each scene had to respond to an autonomous car’s movements, even though the camera may not have captured all the movements. Any image that was not captured by photo or video was simulated using prediction methods, which relied heavily on computer graphics and physics-based prediction techniques. The research team overcame this challenge by developing technology that pinpointed the components of a street scene and presented them in a way that can be resynthesized to create a wide range of photo-realistic driving situations.
AADS can move vehicles and pedestrians from one environment to another, using the right lighting and movement patterns. It can also create different viewing angles of streets to provide improved perspective when a vehicle is moving or changing lanes. Advanced image processing technology also provides smoother transitions and reduced distortions, compared to video simulation techniques that are currently used. Researchers hope companies that develop self-driving technology will begin to incorporate this data-driven approach to create safer driving simulators.
Baltimore Car Accident Lawyers at LeViness, Tolzman & Hamilton Represent Victims of Self-Driving Car Accidents
If you were seriously injured in an accident involving a self-driving car, you are urged to contact the Baltimore car accident lawyers at LeViness, Tolzman & Hamilton as soon as possible. We will thoroughly review the details of your case and walk you through the claims process. We will protect your rights and secure the financial compensation you deserve for your injuries. To schedule a free consultation, call us today at 800-547-4LAW (4529) or contact us online.
Our offices are located in Baltimore, Columbia, Glen Burnie, and Towson, allowing us to represent car accident victims in Maryland, including those in Anne Arundel County, Baltimore County, Carroll County, Harford County, Howard County, Montgomery County, Maryland’s Western Counties, Prince George’s County, Queen Anne’s County, Southern Maryland, and the Eastern Shore, as well as the communities of Catonsville, Essex, Halethorpe, Middle River, Rosedale, Gwynn Oak, Brooklandville, Dundalk, Pikesville, Nottingham, Windsor Mill, Lutherville, Timonium, Sparrows Point, Ridgewood, and Elkridge.