It seems an $880-million-dollar project is underway to implement ECTS across the Sydney Trains network, and bring with it ATO-style operation...From the media release HERE :

Digital Systems to transform Sydney's rail network
Published10 Dec 2018
The transformation of Sydney’s rail network through Digital Systems is a step closer with the NSW Government signing a $16 million contract to ensure new technology is properly integrated into the Sydney Trains network.

Network Rail Consulting has been awarded the System Integrator contract and will play a key role in defining and delivering the Digital Systems program. The System Integrator brings experience from deployment of comparable systems delivered on Thameslink and on the Madrid suburban network, as well as many other projects through Europe and globally.

In June 2018, the NSW Government announced an $880 million investment in technology improvements to modernise the Sydney Trains network, including the development of Digital Systems.

The Digital Systems Program is the next phase of More Trains, More Services, and contains three main elements:

• Replacing trackside signalling equipment with the latest European Train Control System Level 2 technology
• Implementing Automatic Train Operation, which will be used to assist drivers – who will still remain in control – and provide faster and more consistent journey times
• Introducing a Traffic Management System for more effective incident management and service regulation across the network.

These elements will transform Sydney’s rail network, creating improved services for our customers.

The System Integrator will lead collaboration with key partners in delivering Digital Systems. They will also assist with procurement of major supplier packages, which will continue in 2019.

Digital Systems will be delivered in stages with services coming online progressively from the early 2020s.

And here was I thinking that the Australian Taxation Office (ATO) was about to take over running the trains after Gladys had sneakily handed them over in lieu of NSW paying Taxes.

Just as well I got that wrong, imagine the line up at the barriers when the ATO demanded proof that you had paid your taxes before letting you in (or out even).

Silly me.

And here was I thinking that the Australian Taxation Office (ATO) was about to take over running the trains after Gladys had sneakily handed them over in lieu of NSW paying Taxes.

Just as well I got that wrong, imagine the line up at the barriers when the ATO demanded proof that you had paid your taxes before letting you in (or out even).

Silly me.

I'll believe it when ticket barriers start accepting iTunes cards as payment.

Are we sure they didn't just mean Automatic Train Protection, which has been tested for some time, and would be consistent with the "Drivers still being in control" line?

Are we sure they didn't just mean Automatic Train Protection, which has been tested for some time, and would be consistent with the "Drivers still being in control" line?

It appears to be both.

The selection of ETCS Level 2 would suggest that the present ATP system is to be replaced. ETCS Level 1 (used in Adelaide) and ETCS Level 2 both incorporate train protection into the system.

The awarding of contracts to Network Rail Consulting and Go-Ahead Group would suggest that the ATO implementation will be similar to that as done on the Thameslink route, where ATO is implemented on only the high-frequency parts of the network and with drivers still performing a number of control functions including:
(a) checking the platform-train safety before departure,
(b) commanding the train to start once cleared,
(c) initiating an emergency brake application if needed,
(d) driving the train manually during degraded working.

Are we sure they didn't just mean Automatic Train Protection, which has been tested for some time, and would be consistent with the "Drivers still being in control" line?

It appears to be both.

The selection of ETCS Level 2 would suggest that the present ATP system is to be replaced. ETCS Level 1 (used in Adelaide) and ETCS Level 2 both incorporate train protection into the system.

The awarding of contracts to Network Rail Consulting and Go-Ahead Group would suggest that the ATO implementation will be similar to that as done on the Thameslink route, where ATO is implemented on only the high-frequency parts of the network and with drivers still performing a number of control functions including:
(a) checking the platform-train safety before departure,
(b) commanding the train to start once cleared,
(c) initiating an emergency brake application if needed,
(d) driving the train manually during degraded working.

It's also similar to what's proposed for Crossrail (Elizabeth Line) in London, where ETCS Level 2 applies through the central core with onboard drivers monitoring the system and driving manually on the outer branches.  The Victoria Tube line, which is ATO, also has a driver/person monitoring operation because of the London commuters' concern about driverless operation.

It's also similar to what's proposed for Crossrail (Elizabeth Line) in London, where ETCS Level 2 applies through the central core with onboard drivers monitoring the system and driving manually on the outer branches.  The Victoria Tube line, which is ATO, also has a driver/person monitoring operation because of the London commuters' concern about driverless operation.

London commuters appear to be fine with the DLR.

There is a driver on board because British rail safety law requires a 'qualified staff member' on board - hence all DLR trains have a CSO on board who is trained to operate the train if required.

Personally, I think driver supervised automatic train operation is MORE dangerous than unattended ATO, as if a human is still in the loop, certain 'corner cases' are then not handled by the system as the 'human' can take over if it goes wrong. The design engineers take certain shortcuts - the project doesn't spend the money solving those last little problems.
If the project is unattended ATO from the start, they go the extra little bit further to ensure fully fail-safe operation of the control system in all cases.

I've been on a 'supervised' ATO metro where you could see into the driver's cab (glass wall). I watched the driver drive manual for one station, then put the train into ATO mode, watch for one stop, then gets a book out of his bag and proceeds to read a book the rest of the trip, only occasionally looking up.
Had the ATO made a major error, the driver wouldn't have had enough 'situational awareness' to make a sensible decision. Those few extra seconds it takes to assess the situation and react could have the train up the rear of another train or into the side of a train due to a misaligned point before he could react.
One of the touted advantages of ATO is running trains closer together - removing the space needed for human reaction time.

We still have train guards and Constance has recently gone on the record as saying he will keep them because of the old sydney trains system and the large gaps at platforms. ATP is what Sydney trains will be getting and not ATO. ATO is what is happening on the new driverless metro.

I would think they should have some sort of automatic train operation equipment on all trains for one-man operation, with two-man operation remaining on manually-operated trains.

One of the touted advantages of ATO is running trains closer together - removing the space needed for human reaction time.

One of the features of ETCS Level 2 is in-cab signalling, regardless of whether or not the driver is human or robot.

This means the driver - regardless of species - *is* able to see around corners and respond appropriately to the aspect signals they cannot yet eyeball.

It's this which allows higher speeds and tighter spacings, not the relative reaction times.

ETCS Level 3, which AFAIK is still vapourware, is moving block signalling, and (AFAIK) that's the key feature of Elizabeth line that will allow tighter headways again.

We still have train guards and Constance has recently gone on the record as saying he will keep them because of the old sydney trains system and the large gaps at platforms. ATP is what Sydney trains will be getting and not ATO. ATO is what is happening on the new driverless metro.

They have quite clearly stated that ATO is the goal.
The original train stop replacement project was ETCS Level 1 limited supervision - replacing the mechanical train stops. The project has suffered serious feature creep over the last couple of years - it's now ETCS Level 2 with radio block. and the ability to go ATO 'in the future'.

One of the touted advantages of ATO is running trains closer together - removing the space needed for human reaction time.

One of the features of ETCS Level 2 is in-cab signalling, regardless of whether or not the driver is human or robot.

This means the driver - regardless of species - *is* able to see around corners and respond appropriately to the aspect signals they cannot yet eyeball.

It's this which allows higher speeds and tighter spacings, not the relative reaction times.

No, it's the removal of the overlaps that are built into the system in case a driver attempts to SPAD that allows closer headway. The ATP will intervene far earlier if the driver doesn't and the train will NOT overrun the signal (in theory).

No signal should be sighted where a driver can not see it in time to react - if they are any like that out there, that's incredibly bad design. In locations where this happens due to geography, repeaters are installed.

ETCS cab signaling has one endgame - replacing the human driver with a box of electronics. It's not worth the cost otherwise. A properly trained and attentive driver is quite good a reading signal aspects from quite some distance away, the only reason to spend the mega-bucks to relay the signal aspects directly to the cab is to do away with the human in the loop.

In cab signalling means no such thing Matthew. In cab signalling means the signals are inside the train for the driver to see instead of outside. In an ATO system no signals are needed period and this is what the new sydney metro will be achieving. ETCS v1, 2 or 3 is an ATP system and not an ATO system. Someone is getting ahead of themselves by mentioning ATO since etcs doesn't have that in it's standards.

Moving the signals to inside the train and removing signals is being done to help defeat track equipment vandalism that is becoming much more common in Sydney.

One of the touted advantages of ATO is running trains closer together - removing the space needed for human reaction time.

One of the features of ETCS Level 2 is in-cab signalling, regardless of whether or not the driver is human or robot.

...

No, it's the removal of the overlaps that are built into the system in case a driver attempts to SPAD that allows closer headway. The ATP will intervene far earlier if the driver doesn't and the train will NOT overrun the signal (in theory).

No signal should be sighted where a driver can not see it in time to react - if they are any like that out there, that's incredibly bad design. In locations where this happens due to geography, repeaters are installed.

You may well be right about the motivation for installation of ATP, but I'd suggest the failure of RailCorp to implement ETCS Level 1 as recommended by the Waterfall disaster inquiries can be attributed to (more or less) the same reason: that it was potentially a stalking horse for single man train crewing.

In-Cab signalling does not allow a reduction in overlaps AFAIK.  But it does allow:
- Higher allowable speeds for ETCS transponder equipped trains where signal visibility is the determinant of the local speed limit (and that's in a lot of places in Sydney)
- Higher allowable speeds by increasing the number of signal aspects (for in-cab only obviously) to allow for appropriate stopping distances over multiple short signal blocks

Both safety and performance should be greatly improved (but I suspect it won't be ).

In cab signalling means no such thing Matthew. In cab signalling means the signals are inside the train for the driver to see instead of outside. In an ATO system no signals are needed period and this is what the new sydney metro will be achieving. ETCS v1, 2 or 3 is an ATP system and not an ATO system. Someone is getting ahead of themselves by mentioning ATO since etcs doesn't have that in it's standards.

Moving the signals to inside the train and removing signals is being done to help defeat track equipment vandalism that is becoming much more common in Sydney.

The press release quite clearly states that the system being implemented, ETCS Level 2, will have the capability of being upgraded to ATO, even though initially it will be ATP.  ATO means what it says, Automatic Train Operation, whether a driver/attendant is present or not.   I think matthewg might know a little bit more about it.

This is a couple of years old, but ATO is definitely on the roadmap.

https://www.transport.nsw.gov.au/sites/default/files/media/documents/2017/PL-S-44000-Signalling-and-Control-Systems-Strategy.pdf

Note that it will always be Grade 2 LOA (Level of Automation) which requires a driver to still be present in the drivers cab since the system isn’t designed for Grade 4 LOA (like Sydney Metro) which requires a closed system with platform screen doors, track violation sensors etc.

It would be like London Underground, where the driver starts the train, the ATO then controls the speed and stops at the next station, but the driver can apply the emergency brakes if required.

Note that it will always be Grade 2 LOA (Level of Automation) which requires a driver to still be present in the drivers cab since the system isn’t designed for Grade 4 LOA (like Sydney Metro) which requires a closed system with platform screen doors, track violation sensors etc.

It would be like London Underground, where the driver starts the train, the ATO then controls the speed and stops at the next station, but the driver can apply the emergency brakes if required.

And I consider this the most dangerous form of automation. The system isn't fully failsafe as they assume a human is supervising. However, said human won't really be paying attention, got to be one of the worlds more boring jobs, being the babysitter for a mostly automated train.
When something does go wrong, precious seconds are lost while the driver/attendant gains 'situational awareness' and takes action. Or they take an inappropriate action in panic.
The airline industry mostly understands this, and despite aircraft having autoland and auto takeoff as well as 'autopilot' cruise control, most takeoffs and landing are done mostly manually - so the crew keep their skill level up - and more importantly, if something does go wrong are full situationally aware. (A flight engineer neighbour once told me if you have a super smooth landing, don't congratulate the pilot, he left the plane in auto-brake


I've been on an ATO train where due to a glass cab wall, I could clearly see the driver reading a book, not looking out for obstructions ahead. Had there been an obstruction, track fault, misaligned switch, there is no way that guy would have noticed in time to stop the train before impact.

I would assume if they pull off the Grade 2 implementation, after a few years the trains will gain LIDAR/RADAR sensing units for obstruction detection and the system will migrate to LOA 4.

Rio Tinto is (or will be) running huge unit ore trains with no driver on board. One of the reasons for the project in the first place was the difficulty of recruiting drivers and one of the main delay factors to trains was lack of crew. They may still have a supervising driver at this stage, but the endgame is clear - no one on board. The sensing technology deployed for that project will make its way back into urban rail sooner or later.

So don't for one minute think the open nature of the suburban network precludes GOA4.

Rio Tinto is (or will be) running huge unit ore trains with no driver on board. One of the reasons for the project in the first place was the difficulty of recruiting drivers and one of the main delay factors to trains was lack of crew. They may still have a supervising driver at this stage, but the endgame is clear - no one on board. The sensing technology deployed for that project will make its way back into urban rail sooner or later.

So don't for one minute think the open nature of the suburban network precludes GOA4.

Not anymore they're not. Talking to a couple mates up there most trains are going out Driverless, the crew set up the train, hand it over to the Train Controller, wave goodbye, and it reappears in a day-and-a -half loaded with ore.

In an ATO system no signals are needed period and this is what the new sydney metro will be achieving.

Except there are signals on Sydney Metro... so that when the trains need to be in manual control, they can be driven. Whether or not that extends to the full Metro system, I do not know. But there are photos that exist of the areas surrounding the maintenance facility with signals.

ATO has big benefits in closing the train spacing through the city circle.  This is needed from the maxed out train density.  Sydney trains did a presentation on some of the issues to RTSA at the end of this year.  
The Up Shore between Waverton and Central has a lot of trouble recovering from even small delays after 3pm on weekdays.  Given the previous variation in performance of some of the signal relays across the bridge and through Wynyard some of the drivers were pulling back further from the signalled gaps allowed.  The networks runs to the slowest train when timetable density is at the highest.  This is the next development from the multi factored project to manage congestion in the city circle stations.

I wonder if the benefits of A.T.O are any greater on trains for one-man operation than on manually operated trains.

I wonder if the benefits of A.T.O are any greater on trains for one-man operation than on manually operated trains.

It pains me to say it, but yes there are - particularly for crewing... Rio Tinto is finding this with their AutoHaul project. Eliminating changeover points brings with it fuel, brake and (significant) time savings - even though doing away with changeovers and switching to barracks working would do similar without a billion-dollar black hole in their budget...

Granted they are a special case, but similarities still apply. With suburban systems, the removal of Driver's means fewer staff, obviously, but also fewer staff to jack up during EBA negotiations, more consistent stopping points, complete adherence (in theory) to signals and speedboards, easier transposition of services if need be and the above-mentioned closer train spacing due to the less-conservative nature of electronics vs a human who fears a pineapple if he screws up.

Okay, so currently, Sydney trains are manually operated with a crew of two, I gather the plan is to go to automatic operation with a crew of one on automatically operating trains.

Okay, so currently, Sydney trains are manually operated with a crew of two, I gather the plan is to go to automatic operation with a crew of one on automatically operating trains.

No there isn't outside of the metro and the transport minister has stated that he will not implement Driver only operation after being shown the benefit of having guards on a train service.

A guard is better for train and passenger protection then a driver alone could ever be on the Sydney system.

A guard is better for train and passenger protection then a driver alone could ever be on the Sydney system.

If the guards would come out of their hidy hole and actually 'guard', but they are not obliged to.

The only passenger facing duty they appear to do is deploy the ramps for a wheelchair user if the station SA isn't waiting for them.

Door opener and closer could be easily automated.

If the metro line manages to run with less than 100% staffed trains and it isn't a blood-bath, the pressure will be on to remove the 'redundant' guards from the conventional network.

Unattended ATO of the conventional network is coming, it may take a little while longer, but it's coming. However, I suspect the guards will be axed before the trains get fully automated and potentially become 'guard only' trains, but that is a possibility - a staff member stays on all trains, but they will be like the CSAs employed on the London DLR - they are in the cabin with the passengers, patrolling the train and being generally highly visible, but not actually being involved that much in the operation. (OK, DLR CSAs do close the train doors at each station, so they do have a train function, and can and do manual drive when there is a problem with the signals or trains).

At any rate, removal of guards is not something for this current government. In the event they manage to get back for another term, I wouldn't see suburban train guards lasting the full term, give them 6-12 months and it will be war. I don't think Labor has the guts to take on the RBTU, but the Libs just might.

We already have DOO trains on the network, that were hastily modified to have guards controls after introduction to service when the government decided that taking on the RBTU wasn't such a good idea after all.