Alternative to instantly applied emergency brakes

[tomuk]

Moderator note: Split from

Underload SPADS & Stadler 745EMU/755BMU

Am coming to the end of driver training course and been tasked with a project! To which end...... I'm looking at underload & Stadler 745/755s, Non technical skills, human factors, SPADs and SPAD mitigation. Stadler 745/755 CABs are comfortable, everything is ergonomically designed so that...

railforums.co.uk

however a mainline train doing 100mph needs someone at the cab at all times even if it was running automatically incase an incident occurred that required the emergency breaks to be applied instantly

With the braking distance of a train being so long what effect do these instantly applied emergency brakes actually serve? Could this function not be replaced with lidar/radar tech for self driving cars?

 

[DerekC]

With the braking distance of a train being so long what effect do these instantly applied emergency brakes actually serve? Could this function not be replaced with lidar/radar tech for self driving cars?

The problem is that on very long sections of straight track a train can't stop within line of sight, and lidar and radar can't see round corners.

 

[tomuk]

The problem is that on very long sections of straight track a train can't stop within line of sight, and lidar and radar can't see round corners.

Neither can a train driver see round corners. In an ETCS and ATO controlled train what benefit does a driver provide? This isn't a slight on drivers who are currently highly skilled but they would be completely underloaded in such a scenario.

 

[dm1]

In general designing a system to operate automatically under normal conditions is much easier than designing one that can operate under all conditions.

For the latter, it is not sufficient to just design it, one must also prove that it is able to function as safely as or safer than a system with a driver, at all times. That means every possible scenario must be analysed and the system must be proven to handle all of them correctly. The only practical way to do this at the moment is to remove as many interfaces with other systems as possible, which is why driverless, fully unstaffed systems are nearly universally fully grade-separated and are fitted with PEDs at stations. Proving to the level required that a train won't crush a drunk passenger leaning onto the side of the train at the platform under unfavourable circumstances is extremely challenging.

A snazzy smart, digital, machine-learning, computer vision, blockchain, edge computing, IoT algorithm that only works 99% of the time (excuse all the meaningless buzz-words) just isn't good enough for safety-critical applications like railways without a human backup. Autonomous cars have yet to receive safety certification without driver supervision, and safety standards for cars are much lower than those for railways as far as risk tolerance is concerned.

 

[tomuk]

In general designing a system to operate automatically under normal conditions is much easier than designing one that can operate under all conditions.

For the latter, it is not sufficient to just design it, one must also prove that it is able to function as safely as or safer than a system with a driver, at all times. That means every possible scenario must be analysed and the system must be proven to handle all of them correctly. The only practical way to do this at the moment is to remove as many interfaces with other systems as possible, which is why driverless, fully unstaffed systems are nearly universally fully grade-separated and are fitted with PEDs at stations. Proving to the level required that a train won't crush a drunk passenger leaning onto the side of the train at the platform under unfavourable circumstances is extremely challenging.

A snazzy smart, digital, machine-learning, computer vision, blockchain, edge computing, IoT algorithm that only works 99% of the time (excuse all the meaningless buzz-words) just isn't good enough for safety-critical applications like railways without a human backup. Autonomous cars have yet to receive safety certification without driver supervision, and safety standards for cars are much lower than those for railways as far as risk tolerance is concerned.

But current systems aren't infallible, take the Kirkby derailment and the death at James Street.

 

[dm1]

They are not. But any system that replaces the current system is expected to be as safe or safer than the current system. That means the bar for removing the driver completely is extremely high, as keeping the driver there will almost always be statistically safer than removing them.

 

[tomuk]

They are not. But any system that replaces the current system is expected to be as safe or safer than the current system. That means the bar for removing the driver completely is extremely high, as keeping the driver there will almost always be statistically safer than removing them.

I believe that is a bold statement. Society seems to believe that the introduction of self driving cars will end all ills on the road. In the much more controlled environment of the railway the retention of a driver to monitor things will maintain the utmost safety.

 

[D365]

Moderator note: Split from

Underload SPADS & Stadler 745EMU/755BMU

Am coming to the end of driver training course and been tasked with a project! To which end...... I'm looking at underload & Stadler 745/755s, Non technical skills, human factors, SPADs and SPAD mitigation. Stadler 745/755 CABs are comfortable, everything is ergonomically designed so that...

railforums.co.uk

With the braking distance of a train being so long what effect do these instantly applied emergency brakes actually serve? Could this function not be replaced with lidar/radar tech for self driving cars?

Without an emergency brake, how do you propose to stop a train in the event of an overspeed, strike or general loss of control.

 

[edwin_m]

They are not. But any system that replaces the current system is expected to be as safe or safer than the current system. That means the bar for removing the driver completely is extremely high, as keeping the driver there will almost always be statistically safer than removing them.

If some form of driverless operation was introduced, over a period of time in service it would probably be shown to be statistically safer than operation by a driver. Because of the rareness of railway accidents, that period would need to be long to be believable.

However, it would be necessary prove to a high standard that this would be so, before any such operation could be introduced. This is extremely difficult, as it relies on judgment about such things as the frequency and nature of those rare events and the comparison of error rates by people and by systems - both difficult to estimate convincingly.

Driverless operation is now standard for new metros which operate in a fully segregated environment including platform screen doors, but is seen as much more of a challenge on other types of railway where that sort of unforeseen incident is more difficult to design out.

 

[tomuk]

Without an emergency brake, how do you propose to stop a train in the event of an overspeed, strike or general loss of control.

I'm not advocating no emergency brakes! What I'm suggesting is in the cab of tomorrow with ETCS, C-DAS and ATO or at least low technical barriers to ATO drivers will be bored out of their minds therefore would not the application of car self driving technologies such as radar and lidar to cope with track incursions allow for the removal of drivers as they would by then have been forced to endure a mind numbing task.

 

[GC class B1]

I'm not advocating no emergency brakes! What I'm suggesting is in the cab of tomorrow with ETCS, C-DAS and ATO or at least low technical barriers to ATO drivers will be bored out of their minds therefore would not the application of car self driving technologies such as radar and lidar to cope with track incursions allow for the removal of drivers as they would by then have been forced to endure a mind numbing task.

There is no such thing as Emergency Brakes on British Railway vehicles. There is only the Service Brake operated by the Drivers Brake controller (or power brake controller) which the driver uses to initiate all brake applications up to Full Service and Emergency. An Emergency Brake application will be made by the safety systems in the event that the driver does not brake normally as required, for example by TPWS activation. The Emergency Brake application occurs slightly more quickly than the Full Service brake application and in the case of vehicles with enhanced emergency at a higher deceleration rate.
To be correct locomotives and some on-track-plant also have a direct brake but this operates through the same brake actuators and rigging as the Service Brake.

 

[43096]

There is no such thing as Emergency Brakes on Railway vehicles. There is only the Service Brake operated by the Drivers Brake controller (or power brake controller) which the driver uses to initiate all brake applications up to Full Service and Emergency. An Emergency Brake application will be made by the safety systems in the event that the driver does not brake normally as required, for example by TPWS activation. The Emergency Brake application occurs slightly more quickly than the Full Service brake application and in the case of vehicles with enhanced emergency at a higher deceleration rate.

I can't decide if you're trying to be clever or not. There is such a thing as an emergency brake - but it is applied using the same equipment as for any other brake application, though; the difference being in speed of application and/or degree of brake application in comparison with a service brake application. In your treatise, you haven't mentioned the emergency brake plunger on the driver's desk.

 

[GC class B1]

I can't decide if you're trying to be clever or not. There is such a thing as an emergency brake - but it is applied using the same equipment as for any other brake application, though; the difference being in speed of application and/or degree of brake application in comparison with a service brake application. In your treatise, you haven't mentioned the emergency brake plunger on the driver's desk.

I am being factual as incorrect terminology causes problems. An Emergency brake application can be initiated by several means including the plungers. The term Emergency Brake is incorrect and its use can be incorrectly interpreted as referring to a separate brake system. I think the post to which I replied may be misunderstood by readers and I am keen to avoid misunderstandings.

 

[tomuk]

I am being factual as incorrect terminology causes problems. An Emergency brake application can be initiated by several means including the plungers. The term Emergency Brake is incorrect and its use can be incorrectly interpreted as referring to a separate brake system. I think the post to which I replied may be misunderstood by readers and I am keen to avoid misunderstandings.

To avoid confusion I'm not referring to any separate or additional emergency brake such as track or magnetic brakes just to an application of the service brake in full or emergency position by the driver in an unexpected situation such as a track incursion.

 

[ComUtoR]

I am being factual as incorrect terminology causes problems. An Emergency brake application can be initiated by several means including the plungers. The term Emergency Brake is incorrect and its use can be incorrectly interpreted as referring to a separate brake system.

What do you mean by a separate brake system ?

On our units we have an Electro-Dynamic brake, a Pneumatic Friction brake, and a Spring applied parking brake. Each is listed in my manual as 'brake systems' Under 'Emergency Braking' it describes two levels of emergency braking.

 

[GC class B1]

What do you mean by a separate brake system ?

On our units we have an Electro-Dynamic brake, a Pneumatic Friction brake, and a Spring applied parking brake. Each is listed in my manual as 'brake systems' Under 'Emergency Braking' it describes two levels of emergency braking.

To explain, multiple units operating on Network Rail have only one brake system. For EMUs this system comprises three parts (sub systems), the dynamic brake also referred to as the blended brake which operates in conjunction with the friction brakes, the friction brake and the parking brake. The Blended brake forms the Service brake. The function of the parking brake is to hold the unit stationary when the brake air system is not charged and the Service brake will not therefore hold the unit. Depending on the particular unit the service brake has either an analogue control which has a graduated control up to the Full Service application or a number of steps which in the case of most non-intercity type units give around 3, 6 or 9% g deceleration. On Multiple in it’s the Power brake controller has a further setting called Emergency. For DMUs with the energise-to-release 3 step control this does not reduce the brake application time orr increase the dece rate but earths the control wires to ensure that the brakes cannot be released by any electrical inputs. I am not as familiar with non-intercity EMUs, however I do know that the Emergency brake setting increases the deceleration rate to about 12% g. From your post I surmise that the two settings of Emengecy brake application are nominally 9% and 12% g.
The title of this post is ‘alternatives to instantly applied emergency brakes’ and I found this a bit confusing. The term ‘instantly applying, could refer to a system that does not have a delay in applying the brakes but this is not the case with any friction brakes. The dynamic brake is almost instant, but an Emergency brake application will require application of the friction brakes. I think the term may be used for the safety system initiating an Emergency brake application of the friction brake, but this is not instant
The use of the term Emergency Brake suggest that there is a separate braking system other than the service brake that is applied instantly. This could be a magnetic track brake as fitted to trams and some European vehicles but these are not fitted to British trains. I hope that this explains my reasoning.

 

[D365]

The title of this post is ‘alternatives to instantly applied emergency brakes’ and I found this a bit confusing. The term ‘instantly applying, could refer to a system that does not have a delay in applying the brakes but this is not the case with any friction brakes. The dynamic brake is almost instant, but an Emergency brake application will require application of the friction brakes. I think the term may be used for the safety system initiating an Emergency brake application of the friction brake, but this is not instant

Yes, that's the correct interpretation. I think what the OP is inferring is the possibility of an additional train protection system (i.e. beyond baseline ETCS) that would have the capability to demand an emergency brake application. The feasibility of an instant brake application would depend entirely on the vehicle.

 

[ComUtoR]

I hope that this explains my reasoning.

It does, thanks.

Additionally, one of our units uses the TMS for the service braking but has a separate 'system' for emergency braking. When it does use emergency braking, granted there is some blending but the main brake on the motor coaches will still use the Electro-dynamic brake and the friction will kick in if the ED brake is not working. Do you still consider that as 'one system' ?

 

[GC class B1]

It does, thanks.

Additionally, one of our units uses the TMS for the service braking but has a separate 'system' for emergency braking. When it does use emergency braking, granted there is some blending but the main brake on the motor coaches will still use the Electro-dynamic brake and the friction will kick in if the ED brake is not working. Do you still consider that as 'one system' ?

Yes. I consider the brake system to be the system that performs all the functions necessary to decelerate the train and to hold a stationary train. i am not familiar with the system you describe however I expect that the TMS will form one part of the brake control command element of the brake system. This brake command will need to be supplemented by the Driver initiated or safety system initiated Emergency application as it is not permitted to have a single point failure that would result in complete loss of brake force on the train. If TMS was the only means of making a brake application and failure of TMS could result in the driver or safety system not initiating a brake application when demanded, then this would not be compliant with the requirements of a brake system as defined in the standards (TSI loco and Pass and the superseding Railway National Technical Specification Notification (NTSN)).
Emergency braking can use both dynamic and friction braking. However dynamic braking cannot be relied on in all circumstances as loss of the traction supply can result in loss of the dynamic brake as occurred with the Caledonian sleeper train brake incident (RAIB report No 05/2020). As you state in your post the friction brake must be available to stop the train if the dynamic brake is not available.