I wish they'd kept conductors on buses but that ship left 3 decades ago. Some of my fondest memories are the conductors on Melbourne's ageing W class wooden trams.
Trains feel like something where drivers and conductors and platform staff are a social good. We're beyond cost at this point, it's about public utility.
I think the combination of driverless trains and automated platform screen doors is the gold-standard for new metro-like systems. The new REM in Montréal is able to have very high frequencies with much lower cost, so transit users should be able to get more transit for the same cost and the municipality isn't worried about the burden of high operating costs for the next century. The platform screen doors are great too, anyone who's taken the Metro enough has had to wait for hours because someone has fallen onto or committed suicide on the track, which feels like such a sad problem to have when the solution exists.
I used to think that was crazy until I started using the metro every day. There will be a hundred people packed by the doors and the few people in the middle of the car don't think to compress a little. Suddenly I wanted those staff to come in and make the ride more equitable!
Also the ratio of driver to passenger especially compared to say a taxi, is negligible so the amortized cost is basically zero. That said, they’re not exactly useful on trains, you don’t chit chat with the man in the armored cab.
> Airlines are trying to squeeze every penny out in a depressing race to the bottom because they couldn't really differentiate themselves and travelers overwhelmingly just pick the cheapest flight
They also removed the roles of radio operator, navigator, flight engineer and third officer without compromising safety or capability [1]. I'd want to see studies. But single pilot + relief crew should be safe enough to start with for long-haul flights.
Single pilot is fine as long as nothing goes wrong. When a problem occurs then workload can get really severe really fast. You need one pilot flying and the other running checklists and troubleshooting. Even if there is a relief pilot onboard they might be sleeping and unable to make it to the cockpit fast enough to help. The higher risk level might be judged acceptable for cargo flights but I think it will be a long time until automation makes it viable for scheduled passenger flights.
> When a problem occurs then workload can get really severe really fast. You need one pilot flying and the other running checklists and troubleshooting
This argument was made for each of those now-obsolete roles. I'd want to see data before concluding there is a benefit to a second person on checklists. In reality, that advantage would only manifest if the captain is inexperienced, self-check systems have failed and the radio/remote ops have failed.
> Even if there is a relief pilot onboard they might be sleeping and unable to make it to the cockpit fast enough to help
Can you name a crash which occured because the FO was in the bathroom?
> Can you name a crash which occured because the FO was in the bathroom?
Germanwings 9525. Can’t discount human factors, especially in an industry where human factors are very frequently the root cause of incidents.
More importantly I can name any number of flights where two pilots became task saturated and flew a plane into the ground. Or flights where a single pilot used bad CRM to force the other pilot to do something stupid, even though they both knew it was stupid. Both of those would be more common without a second pilot, general aviation is proof of that.
> Germanwings 9525. Can’t discount human factors, especially in an industry where human factors are very frequently the root cause of incidents.
Not a great argument for a second pilot unless the second pilot never goes to the bathroom. The human factors involved in Germanwings 9525 didn't randomly occur during the flight, they specifically occured while the co-pilot was in the bathroom.
> Both of those would be more common without a second pilot, general aviation is proof of that
Between the reliability of piston engines versus jets to the experience of someone with a PPL versus ATP, this comparison includes far more variables than number of pilots.
> I can name any number of flights where two pilots became task saturated and flew a plane into the ground
I can do that for any value of N. The question is what the ratio is between N = 1 and N = 2, and how that has changed with evolving automation.
> where a single pilot used bad CRM to force the other pilot to do something stupid, even though they both knew it was stupid
Not really an argument for a second pilot either...
I'm not saying we have the evidence to say a single pilot is sufficient. I'm arguing that nothing you've brought up shows it's critical. That said, I'm sure we'll wait until another aviation authority pioneers the way. American aviation ceded the first-mover position decades ago.
> Not a great argument for a second pilot unless the second pilot never goes to the bathroom. The human factors involved in Germanwings 9525 didn't randomly occur during the flight, they specifically occured while the co-pilot was in the bathroom.
One of the changes from that incident is that there is always a second person in the cockpit. If a pilot needs to leave the cockpit for some reason, a second crew member must come in and ensure that the pilot flying doesn’t lock the other pilot out.
As to the rest of the argument: the whole point is that two pilots is safer than one by a very large margin. We eliminated engineers, radio operators and navigators because we know have systems that do that better than a human. We do not yet have systems that do the job of pilot as well as a human. Commercial autolanding has existed since the 1960s but there are still situations where the pilot has to land the plane. We aren’t there yet.
> Not a great argument for a second pilot unless the second pilot never goes to the bathroom. The human factors involved in Germanwings 9525 didn't randomly occur during the flight, they specifically occured while the []pilot was in the bathroom.
While that's true, it is a good argument for a third pilot.
To stop one crash in decades amidst hundreds of thousands of flights per day [1]? If you place that high of a value on a human life, there are better optimisations to do first at much lower cost.
That's a different argument about the third pilot.
I'm just pointing out that while my parent was correct that the argument failed as a justification for two pilots, it doesn't fail for any other number. It's similar to the proof that all horses are the same color.
> while my parent was correct that the argument failed as a justification for two pilots, it doesn't fail for any other number
It doesn't fail as much as not work. One can always come up with hypotheticals justifying more redundancy. Whether that's worth the tradeoff requires actually looking at the facts.
> Can you name a crash which occured because the FO was in the bathroom?
I don't think that's a strong argument. The only time a pilot goes to the bathroom is during the most boring portion of the flight when everything is going exactly as expected. Planes so rarely have excitement occur at that phase of flight that the odds are really low of it happening at the exact moment one of the pilots is using the head.
You’d think they were already investigating the use of automation in emergencies. Like going through checklists, debugging situations in real time, etc…it seems like an area where AI could actually be useful beyond current auto pilot/landing tech.
It's just a bunch of if conditions, no need for AI. The problem is that sometimes something not planned in those if conditions happens, and then you need deduction and high knowledge of physics and the aircraft systems.
On modern planes there is a very high degree of automation already. But it's a fine balance, because too complex automations can be confusing or unpredictable to the pilots. There are a few crashes due to pilots not understanding/fighting with automation.
Airlines are trying to squeeze every penny out in a depressing race to the bottom because they couldn't really differentiate themselves and travelers overwhelmingly just pick the cheapest flight.
Public transit is quite different. It's not as opex-intensive as flying, it's usually subsidized by the state, and it's often pretty much a monopoly.
The numbers are pretty different, though. A subway driver might make $50K/yr and the train has a thousand people, with short runs at high frequency. A long haul flight will have 300-400 passengers for hours and 3-4 pilots each making $250K/yr.
It's a bit more lopsided in the US. A regional carrier pilot makes peanuts (due to how they are paid, only when wheels are up, there have been pilots effectively paid lower than minimum wage). At regular carrier like American/United/Alaska, pilots make waaaaay more. A captain may make half a mil. If you can suffer through the regionals and make your way up the ranks, it eventually pays off.
The conductor wages over the system as a whole are a large part of operating costs. Another issue you can have is a lack of conductors, which is a big issue in the Netherlands apparently.
More to the point, staffing isn't setting the price of transport, it's subsidised almost everywhere. Sure. It's a cost. The cost:benefit here is more than just apparent budget impact because public transport is a utility function.
I don't disagree. One thing that RMTransit on Youtube talks a lot about is that transit is a lot more expensive to build in the Anglosphere of the US, Canada, Britain, and Australia, and so despite these countries investing a lot of money into it recently, there's an order of magnitude less return on investment than similar projects happening in the rest of Europe and Asia.
Return on investment not being profit for the transit system, but moreso a more extensive, faster network for riders.
I used scotrail for the first time in years as a casual return visitor to the UK from Oz. It was ok and on time and clean, and the QR code on/off experience was fine but the confusion of choices trying to book.. complete nightmare. Privatisation has only partially been unwound and multi app multi provider timetable and ticket exposure to casual users is away with the bees.
if you look at the accounts/statistics: staff is the majority of the cost
the official figures here are however misleaading as their "staff" numbers only include those directly employed by TOCs and Network Rail
under "other" there's the vast army of contractors, including cleaning staff, track maintenance contractors, train maintenance contractors, private security, and so on
what's left is fuel, interest payments, depreciation, premium payments to government and some capital expenditure (not much...)
Trains feel like something where drivers and conductors and platform staff are a social good. We're beyond cost at this point, it's about public utility.