3 ms·
I was thinking about this for a second. Specifically about saving on employees. I guess it is safe to assume on-board jobs don't have a bright future. Those ar
by smhg 10y ago
I was thinking about this for a second. Specifically about saving on employees.
I guess it is safe to assume on-board jobs don't have a bright future. Those are up for automation.
So maintenance staff remains? But can't that be a separate entity? Just like with cars? That might solve the safety issue too. One has a bad rep? You go to another.
It's just that one public operator doesn't seem to work well either (lots of issues in countries with those too).
- Symbiote 10y agoTrack and signal maintenance is a separate, public entity in the UK ("Network Rail"), with trains owned by the banks, and leased to companies that run services, like Virgin (yes, that Virgin) or Southern. In the 1990s, the track and signals were maintained by Railtrack, a private company, but that caused several bad accidents. Since Network Rail took over, the British railway network has become one of the safest in the world -- I don't think a passenger on a train has been killed for a decade, and only a very small number (2?) of maintenance staff.
- smhg 10y agoYes, it sounds hard to create competition for track and signal maintenance. But the parent was talking about the companies that run the trains. I meant the maintenance those trains need (repairs/safety/cleaning/...). On that level competition might make more sense (like with cars). So they wouldn't do that in-house. Maybe it already works like this?
- Symbiote 10y agoIn Britain, that's all done privately. I think the only exception is the maintenance of London Underground trains, and possibly cleaning. eVen the other publicly owned railways in London contract out cleaning and maintenance.
- crdoconnor 10y ago>I guess it is safe to assume on-board jobs don't have a bright future. Those are up for automation. If neoliberals really believed the "magic robots are coming to steal everybody's job" story they're pushing they'd not have a problem with raising public spending. Automation is inherently deflationary and the more of a 'danger' it is to jobs the larger the deflationary impact it will have. You'll notice that they're still scaremongering about inflation, recommending against increasing infrastructure spending and blaming the lack of jobs on magic robots rather than austerity.