Drive.AI Launching Service in Frisco, TX

[Rockster]

Drive.AI is launching an autonomous ride-hailing service in Frisco, TX, between Hall Office Park and The Star (Cowboy's training complex).

I signed up to be a test passenger tomorrow, the first day of public testing.

 

[TaoJones]

Drive.AI is launching an autonomous ride-hailing service in Frisco, TX, between Hall Office Park and The Star (Cowboy's training complex).

I signed up to be a test passenger tomorrow, the first day of public testing.

What a shame. We hardly knew ye.

Kidding - Looking forward to hearing how it goes and how anomalies are managed.

 

[Cosmacelf]

Let us know how it goes. This is the very first AI car company (other than Waymo) that seems to know what it is doing. They are testing within a restricted area. They have external communications on the van for pedestrians. They make it obvious this is a self driving van (again for pedestrian safety). Once they go away from needing safety drivers, they’ll still have real time remote monitoring personnel. Unlike Uber, they aren’t pushing out barely ready software with inattentive “safety” drivers...

 

[Rockster]

The ride was brief and uneventful. We departed from an office building parking lot, drove through the office complex, crossed a four lane divided road into another complex, and stopped at a predetermined location to swap passengers. The van seemed to go no faster than 10 or 11 mph the entire time. We were permitted to take photographs but no video.

Some observations:

• The system is clearly designed to be extremely cautious. For example, while traveling through a parking lot, the van stopped for nearly every turn, even in the absence of cross traffic or a stop sign.
  
• The van accelerated and decelerated very gently.
  
• I detected some very slight wiggle within the lane, though overall the van stayed well centered.
• While crossing the four lane, divided road, the system successfully accommodated a slight offset in our lanes of travel.
• Understand, too, that this autonomous van has been traveling the office park for weeks in advance of yesterdays public test drive. It's clear that their deployment protocol is to extensively map a discrete area and then deploy into that area. (Not a criticism, just an observation.) I'm interested to know how the system would react to the overnight appearance of, say, a closed lane or a stalled vehicle or some other alteration to its pre-mapped area. Our brief ride didn't afford the opportunity to delve into such details.