If trains are completely separated from pedestrians or any other form of transportation, then 100% automation was possible a century ago. As soon as they’re not, the problem becomes much, much harder, and is not fully solved yet.
Trains take so long to stop that having a human in charge isn’t going to do any good in terms of avoiding pedestrians or any other form of transportation.
However, since you only need one person driving a train that can carry hundreds of passengers it doesn’t really make sense to replace them with AI. The small savings of having one less salary to pay is offset by the fact that you no longer have someone to blame when things go wrong.
I should clarify. When I said “all other forms of transportation”, that includes other trains. You can easily do it on a straight shot running back and forth. Which is what a lot of large airport trams do, and the monorail in Seattle (which is more like a bus on a concrete rail), but not much else.
As soon as you enter the complication of signaling, automation becomes a lot harder.
But yes, it’s not worthwhile to automate away that one extra job.
Self driving trains tho. Not all trains do that
Why? The one driver per hundreds of passengers is drop in the bucket and it is better to have a human who can deal with various issues an AI can’t.
If trains are completely separated from pedestrians or any other form of transportation, then 100% automation was possible a century ago. As soon as they’re not, the problem becomes much, much harder, and is not fully solved yet.
Trains take so long to stop that having a human in charge isn’t going to do any good in terms of avoiding pedestrians or any other form of transportation.
However, since you only need one person driving a train that can carry hundreds of passengers it doesn’t really make sense to replace them with AI. The small savings of having one less salary to pay is offset by the fact that you no longer have someone to blame when things go wrong.
I should clarify. When I said “all other forms of transportation”, that includes other trains. You can easily do it on a straight shot running back and forth. Which is what a lot of large airport trams do, and the monorail in Seattle (which is more like a bus on a concrete rail), but not much else.
As soon as you enter the complication of signaling, automation becomes a lot harder.
But yes, it’s not worthwhile to automate away that one extra job.