Four alternative paths for JTA's Skyway U2C proposal

[thelakelander]

Yes, JTA is shooting for Level 4. Basically the pilot Navya just launched in France:

https://www.auvsi.org/industry-news/navya-launches-level-4-fully-autonomous-shuttle-service-france

[Charles Hunter]

From the article Lake linked to

The shuttle is capable of operating without a driver in a determined environment and on a predefined route thanks to the continuous improvement of the "Navya Driver" autonomous driving software and the sensor architecture of the Autonom Shuttle Evo. If necessary, an off-board supervision can take control of the vehicle in real time. The presence of other users—pedestrians, cyclists and other vehicles—on the site is regulated thanks to secure access and reduced driving speed.

Will that be one remote operator per vehicle?  If so, where is the labor savings?  If not, how many AVs can the "off-board supervision" supervise safely? 
Also, "The presence of other users—pedestrians, cyclists and other vehicles—on the site is regulated thanks to secure access and reduced driving speed."  The AVs are traveling about 11 mph.  How does all of this relate to Bay Street in mixed traffic - unregulated - and, during events, lots of pedestrians crossing wherever?

[marcuscnelson]

Perhaps, but you have to get through one to get to the other, right? Right now most of the industry seems to be just reaching some degree of confidence in Level 3. Some degree of Level 4 has to be met before work can be done on Level 5, right? A pilot isn't going to be the same thing as being production-ready for mixed traffic capability.

JTA's listed requirements include bidirectional operation at 35mph. That alone seems pretty daunting right now.

[thelakelander]

There may be some vehicles that can get up to 35 mph but they'll likely still be regulated closer to the 15 mph range. That's something JTA and their consulting teams have no control over. They did tell me that they expect that they (JTA) will be at Level 4 automation at their test facility within a year or so.

[jaxlongtimer]

From the article Lake linked to

The shuttle is capable of operating without a driver in a determined environment and on a predefined route thanks to the continuous improvement of the "Navya Driver" autonomous driving software and the sensor architecture of the Autonom Shuttle Evo. If necessary, an off-board supervision can take control of the vehicle in real time. The presence of other users—pedestrians, cyclists and other vehicles—on the site is regulated thanks to secure access and reduced driving speed.

Will that be one remote operator per vehicle?  If so, where is the labor savings?  If not, how many AVs can the "off-board supervision" supervise safely? 
Also, "The presence of other users—pedestrians, cyclists and other vehicles—on the site is regulated thanks to secure access and reduced driving speed."  The AVs are traveling about 11 mph.  How does all of this relate to Bay Street in mixed traffic - unregulated - and, during events, lots of pedestrians crossing wherever?

Charles, I raised these same issues in a previous post and am still waiting for an answer from someone that makes operational and economic sense.  If there is a "driver," whether in the vehicle or remotely, that must be fully engaged in watching the vehicle, then what have we accomplished over a conventional vehicle?  Maybe even going backward as a remote driver probably doesn't have the full sense of place that a driver in the vehicle would. 

Add, as you note, vehicles are traveling at 15 mph or less in mixed traffic.  Aside from the bottleneck an AV will cause, what if there is a traffic bottleneck at grade - how does the vehicle maintain scheduled headways and/or can all the AV's end up bunched together?

And with all due respect to others here, Level 4 or 5, JTA seems a long ways from where they will need to be to make this work in the real world.  As you look at the several articles I have recently posted about AV's and efforts by others, it is pretty unanimous that no one has a desirable result figured out currently, is meeting any reasonable timeline laid out originally, is on budget or can even begin to predict when they will have it figured out.  Existing setups are under controlled, supervised and/or very limited conditions with very cautious settings/configurations.  Far from a full real world application.

The real issue here is why should Jacksonville taxpayers pay for the R & D to solve AV issues?  Are we going to own the "proven" (if ever) technology and collect licensing fees?  I doubt it.  So, what's in it for us other than blowing a lot of money that either is for a failed demonstration (echos of the existing Skyway) or a (most unlikely) success that a for-profit partner gets to fully exploit at taxpayer expense.

I might add that even if we were to successfully prove the technology, I doubt there will be demand that justifies the investment.  So why even pursue it?