My opinion is that the main advantage individual owners using their Teslas as RoboTaxis is where the cars are located. (e.g. all over the place) This has the potential of being able to pick someone up within a very short time period (as opposed to Uber which sometimes never comes at all). Any centrally located facility means the pickup time is typically the distance between the facility and you, although sometimes you can get lucky and the car will be close by. Anyway, what I recall from Investor Day was that Tesla would only supplement individual ownership of RTs to fill gaps.
One of the coolest things about having autonomous Robotaxis is that you could predict demand in an urban area and pre place the fleet to have the least wait time for passenger, or whatever metric you would like to optimise for. Many ride hailers have such service, however their drivers will never be as obedient as their AV counterparts. They have many excuses like "ah well I have been driving for X years, I know where to go to get rides" or I don't want to go that part of the town due to congestion etc . In the case of AV's, you could predict and optimise and make the whole network super super optimised. We had models that would show 50%-75% reduction in Passenger pickup times (ofcourse it was only simulations with many assumptions)
At the boom and hype of AV's in 2017 and 2018, many startups like autofleet, fleetonomy, rideOS etc. focused on solving exactly that.
Source: I worked on optimization problems for fleet balancing for autonomous vehicle for a OEM and the UBER of Europe.
Last edited: