At the 2016 Consumer Electronics Show (CES) in January, Audi presented its new E-tron Quattro concept car. As a 100-percent electric vehicle with a 310-mile range between charges, not to mention the latest in autonomous driving technology, the E-tron Quattro is a fairly accurate representation of how Audi’s first fully-autonomous car will look — and drive.
After some quality time spent with Audi’s innovators and Google’s self-driving car I’ve assembled 10 critical points related to the autonomous future we’re rapidly speeding toward. Many of these items were news to me, and all of them helped me understand the challenges facing autonomous vehicle technology.
- 6 Levels of Autonomous Driving: The National Highway Traffic Safety Administration (NHTSA) has defined six levels of autonomous driving.
• Level 0 is what you’d expect – zero driver assistance from the car.
• Level 1 indicates one or more driver assistance systems, like stability control and/or cruise control, but the systems are not linked to each other.
• Level 2 is two or more linked functions, like lane-keeping assist and collision/brake mitigation working together, but the driver is still expected to be alert and ready to instantly seize control of the vehicle.
• Level 3 combines several automated functions and allows the car to be primarily in control for extended periods of time. This is the first level where transitions from car control to human control can be more relaxed, taking up to 10 seconds and allowing the driver to be fairly oblivious to the driving process while the car is in control.
• Level 4 vehicle is similar to Level 3 but the car now has the ability to find its own way out of danger (i.e. exit the highway or route through heavy traffic to find a safe place to stop) if the driver isn’t responding.
• At Level 5 an autonomous vehicle can perform all driving functions with no human input needed at any time. This means Level 5 cars can operate without anyone inside. - Legal Challenges: Currently it is legal in every U.S. state for cars to operate up to Level 3, except in New York.
- Mode Confusion: Studies show that when a human is fully disengaged from driving it takes at least 10 seconds to get them fully re-engaged, which can be a big problem if an emergency arises in which a computer-controlled car needs to revert to human control. Most automotive emergencies don’t include a 10-second transition period to make sure the driver is ready to face them.
- Hyper-Accurate Mapping: When mapping technology becomes hyper-accurate the car will know when its human driver should take control, and will alert the driver at least 10 seconds before entering regions that require human attention.
- Open Mapping Platform: Establishing and maintaining this level of hyper-accurate mapping will be difficult, but if an open/standard platform for mapping is established across all automakers the process will happen much quicker and be far easier to maintain.
Read more of the original article in Forbes.