Deadly Test Drive (02:40)
On March 18, 2018, Uber's self-driving Volvo hits a Tempe, Arizona cyclist, killing Elaine Herzberg. (Credits)
Why Have Automated Cars? (02:54)
Many Americans say they are afraid to ride in a self-driving car. Advocates state that autonomous cars will improve safety, reduce pollution, eliminate the need to own a car, and provide greater mobility for those who cannot drive.
Driving Safety (03:24)
Zoox founder Jesse Levinson demonstrates autonomous navigation in San Francisco. Despite the complexity of driving, only one accident occurs in 100 million miles of driving per year.
DARPA Grand Challenge (05:55)
The Defense Advanced Research Projects Agency organizes competitions for driverless cars, triggering a race to build a new industry. Many believe the market will be worth trillions by the 2030s.
Uber Crash Investigation (03:56)
Uber believes eliminating paid drivers is important to future profitability. After the March 2018 accident, Uber suspends testing on public roads and the NTSB launches an investigation. Vasquez denies watching a show while in the car.
Levels of Automation (05:48)
Experts classify automation into levels, with five representing the highest level of computer control. MIT researchers study how drivers use partially automated cars. Tesla founder Elon Musk has long believed autopilot is a step toward full autonomy.
Early Adopter Deaths (03:26)
Joshua Brown makes YouTube videos showcasing his Tesla on autopilot; the company provides usage recommendations. Brown dies on May 7, 2016 after hitting a semi-trailer at 74 miles per hour while using autopilot.
False Sense of Security (02:44)
Tesla states that drivers using autopilot have lower a crash rate than drivers as a whole in the U.S. Fully manual automobiles require drivers to use all cognitive resources; automation results in driver inattention.
Drivings Tasks (04:50)
Fully automated cars must be able to see what is around the car, understand the visual input, and plan the car's path while controlling acceleration, braking and steering. Autonomous cars rely on cameras, radar, and LIDAR; each has areas of weakness.
"Edge Cases" (02:38)
Phil Koopman’s company looks for unexpected ways perception software can fail and uses the information to improve training software. Autonomous cars have difficulty with things they have not seen before.
Autonomous Planning (05:18)
Software must anticipate scenarios, determine speed and pathway, and make immediate adjustments if sensors detect trouble. Most companies train planning software in controlled environments.
Closing the Skill Gap (02:10)
Engineers improve autonomous abilities by exploiting tasks computers do well, like precise machine control. A Stanford University team is developing software that allows automated vehicles to take evasive action in an emergency.
Trusting Self-Driving Cars (04:55)
Michael Fleming sees potential conflict between automated cars and human drivers. These include ruler breakers, aggressive drivers, job loss, and ethics. Torc Robotics uses data gathered from encounters on the road to improve software.
Credits: Look Who's Driving (00:46)
Credits: Look Who's Driving
For additional digital leasing and purchase options contact a media consultant at 800-257-5126
(press option 3) or firstname.lastname@example.org.