Wrong Way Driver! Future Tech Solution: Autonomous Vehicles
Earlier this week, the need for autonomous vehicles was reinforced when my brother-in-law was met head-on by an elderly driver going the wrong way on the Corona Express Way in Chino, Ca. He captured the incident on his dash cam, which he posted on social media. The video went viral on Facebook in a matter of hours. Here it is on YouTube.
He exclaimed, "it was one of the most scary moments of my life..." After reviewing the footage, he noticed two California Highway Patrol motorcycles on the other side of the freeway. Later it was learned that the driver was stopped and no one was hurt.
As of 2017, there are many semi-autonomous vehicles on the road. All require human drivers who can take control if the vehicle does something unexpected. This brings a whole list of other issues that can arise, but that is for another topic. What I want to address in this post is what I think it will take for fully autonomous vehicles to become a reality. And, by autonomous vehicles, I mean SAE Level Five - Full Autonomous.
In late 2016, SAE International announced its Levels of Automation for Defining Driving Automation in On-Road Motor Vehicles was adopted by the U.S. Department of Transportation. This defines six levels, from level zero with no automation to level five with full automation.
Here is what I think will be needed:
- Compute Power - Think server processing power
- Sensor suite - LiDAR, radar, ultrasonic, image
- GPS - Accurate GPS that can tell which lane you're traveling in
- Connectivity - Internet, Cloud car-to-car
For the computer you will need better than server firmware. This means, while you need server processing power, your autonomous vehicle's computer will need advanced BIOS features, such as:
- Fast boot times of under half a second (less than 500 ms)
- Predictive failure for hardware components
- BMC Baseboard management controllers to provide out of band monitoring of system
- Secure boot and secure over-the-air (OTA) firmware updates
The onboard computer will need to be on a platform that has a minimum product lifecycle of 15 years. Security updates would have to be provided beyond that point for the expected life of the vehicle.
If you are interested in finding out how you can become involved in the field of autonomous vehicles, a good place to start is looking at current technology companies. I believe that is where some of the core technological innovations will come from.