Sure, my point is that as rare as they are, there are human error accidents that could be avoided if we switched to self-driving trains (and they proved safe). They are already in use at airports and commuter lines.
Problem is, what is the tradeoff. So you switch to driverless trains. First of all, it has only worked in completely isolated systems until now (no mainline railways). Then, you have the problem that when something goes wrong, the consequences in terms of delays are much larger, often making the news, because there’s no one there that can just “quickly do something” to get around the problem and keep going. People tend to get stuck in trains for long times when issues arise - see the REM in Montreal. What’s the consequence of unreliable public transit? People tend to avoid it more, often switching back to their cars for important trips - a much more dangerous form of transportation.
It’s not even a “what if”. And besides, there’s a saying in the railroad: the safest train is a stopped train. In other words, as soon as you start moving a train, you increase the level of danger, no matter what you do. That’s a tradeoff we make as a society. And that tradeoff is much higher in cars than in trains (hence, why public transit can legitimately be seen as a lifesaving measure).
How much a driver can do if a train breaks down depends on the train and how much the driver knows about his engine. Unfortunately, with a lot of the engines being computer controlled, the amount a driver can do has reduced - but there are still plenty of things a driver can repair. And sometimes it’s just a matter of bypassing problems. For example, there’s a problem on the track and you need to give the driver explicit instructions on how to proceed and what to watch for. That’s not something you can easily tell a self driving train to do - that train will most likely be stuck until someone remotes in. Assuming, of course, that the remote connection isn’t shot. Then you’re absolutely screwed.
And even if the train breaks down completely, it’s still important to have someone there so that the passengers aren’t left alone. Who can help evacuate the train if necessary. Who can communicate to the passengers and be there directly if there are specific problems that need to be addressed.
Sure, my point is that as rare as they are, there are human error accidents that could be avoided if we switched to self-driving trains (and they proved safe). They are already in use at airports and commuter lines.
Problem is, what is the tradeoff. So you switch to driverless trains. First of all, it has only worked in completely isolated systems until now (no mainline railways). Then, you have the problem that when something goes wrong, the consequences in terms of delays are much larger, often making the news, because there’s no one there that can just “quickly do something” to get around the problem and keep going. People tend to get stuck in trains for long times when issues arise - see the REM in Montreal. What’s the consequence of unreliable public transit? People tend to avoid it more, often switching back to their cars for important trips - a much more dangerous form of transportation.
So how many lives is this “what if” worth? How much can a driver “quickly do” if a train breaks down, anyway?
It’s not even a “what if”. And besides, there’s a saying in the railroad: the safest train is a stopped train. In other words, as soon as you start moving a train, you increase the level of danger, no matter what you do. That’s a tradeoff we make as a society. And that tradeoff is much higher in cars than in trains (hence, why public transit can legitimately be seen as a lifesaving measure).
How much a driver can do if a train breaks down depends on the train and how much the driver knows about his engine. Unfortunately, with a lot of the engines being computer controlled, the amount a driver can do has reduced - but there are still plenty of things a driver can repair. And sometimes it’s just a matter of bypassing problems. For example, there’s a problem on the track and you need to give the driver explicit instructions on how to proceed and what to watch for. That’s not something you can easily tell a self driving train to do - that train will most likely be stuck until someone remotes in. Assuming, of course, that the remote connection isn’t shot. Then you’re absolutely screwed.
And even if the train breaks down completely, it’s still important to have someone there so that the passengers aren’t left alone. Who can help evacuate the train if necessary. Who can communicate to the passengers and be there directly if there are specific problems that need to be addressed.
Those same computer repairs can be done remotely. Or you have a mechanic on board for repairs, just not driving the train.
Besides, trains already have a dozen crew anyways, serving drinks, checking tickets, etc. No need to risk lives when we don’t need to.