Six months ago, I wrote that “the REAL name of self-driving cars must become something like SOMT: Shared, On-Demand, Micro… TRAIN”. Today, I realized I should explain better a part of that concept, because I received on Twitter the following, sensible critique:

Well, by definition a train runs in predetermined courses. I find it hard to imagine how it could work in practice to have a transport that is both separated by pedestrians etc and does not run in predetermined courses”.

First of all, I need to explain that whenever I say “train” in this context, I ONLY mean “train” as in “runs inside its own spaces, that is set of roads, that are PHYSICALLY separated by those of pedestrians”, like the all the tracks of a subway network are separated by its platforms and stairs. I do not mean AT ALL “train” as something that can only follow “PREDETERMINED courses”.

I firs tried to explain this with words in the “Self-driving cars, you’re calling them wrong” part of my original post. Now, I’ll try with a picture. If I could draw I’d be rich, but even this UGLY thing:

This is what I mean by "treating driverless cars like trains" /img/shared-on-demand-micro-train-concept-small.jpg
click for larger version

should show what I mean by “separated transport without predetermined course”, etc: add fences, overpasses, underpasses, dedicated stops… so that the roads can be full of any mix of driverless vehicles (cars, buses…), the walkways can be much larger than today, and in any case the only points of physical contacts between them and pedestrians are the “stops” in the picture. Basically, I believe that:

  1. it is much simpler to adapt cities in this way, than waste billions chasing software, Artificial Intelligence… good enough to avoid accidents like Tempe (which also prove how dumb the whole idea of “driverless car with emergency driver is). For the record, even some car executives seem WELL aware that stuff like I suggest (“major overhaul of cities”) is much more doable/sensible than any hope to get driverless cars really able to cruise today’s cities
  2. that approach also makes a huge, sorely needed difference as far as jobs and social stability are concerned. Driverless cars with artificial intelligence so sophisticated to drive in the cities of today would create very few, extremely skilled jobs, far far away. Fences etc as I suggest, instead, would create lots of LOCAL, non-outsourceable jobs accessible to unskilled people.

The second point is better explained in the “SOMTs creates JOBS. Lots of REAL jobs” part of my original post.