author photo
By Bruce Sussman
Tue | Jul 31, 2018 | 8:05 AM PDT

After a pedestrian was hit and killed by a driverless vehicle in March 2018, we suddenly heard a lot less about driverless AI projects hitting the road.

Now that is changing again.

One example is a new round of AI-powered driverless vehicles on the road in the town of Frisco, Texas, north of Dallas. They are looking to improve self-driving vehicles and their damaged reputation.

The vans shuttle passengers within a geofenced part of town, and customers hail the self-driving ride with an app.

Here are four things that are different about these vehicles, which are powered by a company called

1. They are the brightest painted self-driving vehicles on the road (so you can't miss them).


2. They have four LED screens on the outside of the van to notify those nearby what, exactly, the van is doing. This replaces gestures a human driver might make to someone.
  • “Waiting,” "Going," "Entering / Exiting,” for when passengers are loading, and "Waiting for You to Cross" which tells pedestrians what the van's intentions are at any point.


3. The driverless vehicles allow passengers inside to see exactly what the vehicle sees while it's driving. Look at this:


4. These driverless vehicles were trained using complicated 3D-mapping of what the company says are real-world scenarios unique to the area where the autonomous ride hailing vans operate.


The vehicles and their artificial intelligence are still learning. And that takes an incredible amount of time because of the data labeling that must be done behind the scenes.

"Annotation is the process of labeling all of the data collected on the road in order for our systems to understand what they are seeing, and ensure they learn how to discern and differentiate stimuli. Annotation is a critical part of the self-driving cycle. For every one hour driven, it takes approximately 800 human hours to label it," posted on its blog.

Of course, as a passenger, you might not think about all the annotation going on. Instead, you'll be focused on getting from place to place, safely, in the hands of artificial intelligence.