Tornado

The major problem of technology is that when one becomes reliant on it and presumes it is infallible thus ignoring the human factor and its value in over-riding faulty decisions.

 

The fact that Network Rail (and its predecessors) have consistently failed in the simple, repetitive task of keeping the cutting sides and cesses clear of greenery is hardly the fault of the signalling system (or any signalling system).
Pat

 

The technology supported good decision making that day. It couldn’t prevent the consequences of other bad decisions about maintenance.

Indeed , one could construct a hypothesis that ETCS, had it been fitted, would have mitigated or even prevented the Salisbury Tunnel Junction collision by removing the scope for the braking point identification error* commented on in the RAIB report.

* - to note, no criticism of the driver is intended. He was driving in the dark, unlit, and in poor conditions, the braking point selected was easily missed.

 

Not to mention the aircraft industry. On Air France Flight 447 from Rio de Janeiro to Paris in June 2009, the pitot tubes on an Airbus A330 froze up due to inadequate heating. Due to this the autopilot disengaged, something the three pilots were not prepared for. To silence the alarms, the airplane was flown at an inclination of about 40deg with the engines at 100%, with the inevitable result that the plane went into an extended stall, eventually crashing into the Atlantic.

https://en.wikipedia.org/wiki/Air_France_Flight_447

 

I'm not sure that's right. I haven't (yet, anyway) had the pleasure of working on an ERTMS project, but I'm familiar with other self-driving signalling systems and my employer sent me to a conference on AI in the rail industry a few months back. There I heard a lot about its potential use for things like timetabling and maintenance scheduling and so on, and nothing whatsoever about signalling systems. Indeed when asked, the presenters seemed pretty clear that AI directly contradicts various standards, which require systems to be predictable and any incidents which arise to be able to be replicated as part of the investigation. If you think about it logically, in a machine learning scenario which goes "oops, that caused a crash, better not do that again" it is going to be impossible to figure out why it happened in the first place.
In time, I could see there being some uses in the more routine design phases, but that might be about it.
And even then, I wonder where the principles engineers of the future will come from, if they can't cut their teeth on some basic circuit design or data early in their careers.

As a complete aside, Artifical Intelligence may be the latest buzz word but while clearing my Dad's house I came across a mid-1980s computing text book which uses the term on the back cover!

 

Which is why I asked what AI means in this context. Machine learning is about more than avoiding the ultimate consequence of a crash - it may simply be applied to help the system adapt to the impact of changing timetables.

 

Well after leaf fall season I experienced an all-wheels-locked skid when braking very tentatively into Stevenage Station Platform 4 - it had rained overnight and then froze on the rails. Class 365 Networkers were generally quite good in adverse conditions but the two four-car units [32 axles or 64 wheels] showed Omph on my speedo and putting the controller into Emergency did not achieve any better retardation. A Cab Secure Radio call to the Kings Cross Signaller, to report the overshoot by about half a coach, elicited that I was not the first that morning and someone had been despatched to try and give the rails more grip. I believe a basic bucket of sand was the end result.
ERTMS might have helped during another Winter's travails however. I was bringing an 8-car 365 from Peterborough to KX in heavy snow. This blizzard caused much disruption, of course, but I reached Hitchin eventually. After the station stop I continued out onto the main line, as signalled, for Stevenage and would then be non-stop to KX. Snow was still being blown from the North so signal hoods were filling up and masking the Up signals to a large extent [Hooray for AWS!]. However, I was stopped at K666 (showing Red) which protected Stevenage Up fast line Platform 2. Once again I used the CSR to call KX signal box. When I reported where my train was standing the signaller did not believe me. I knew one of my KX driver colleagues was behind me, going in to work, so I asked him into my cab to speak with the signaller to confirm our position.
Apparently, the KX signal box display still showed my train back at Hitchin Up platform - so I had travelled all the way to Stevenage without tripping any of the many track circuits! This situation worried me immensely until I got confirmation that all controlled signals were at danger behind my train. After getting clearance to proceed I had to collect a PW gang at Stevenage, take them to nearby Langley Junction to unfreeze point work, after which the snow fall became progressively lighter with signals clearly visible and track circuits working properly. By KX it was clear that no snow at all had fallen but incoming trains like mine were witness to the conditions further north. What a night that was.

 

Latest email just landed, railtour with Vintage Trains on December 10th to York and more at planning stage for next year - meeting clearly was productive last month!

 

There's some confusion on the Tornado Telegraph as to whether the 10th December is aWednesday or Saturday, though VT's booking page clearly has it as Wednesday the 10th December... good to see whichever !!!

 

Reading Michael Whitehouse's update on the Vintage Trains thread, Clun Castle is due to be retubed and could be out for some of next year. 5043 is being used sparingly as they are trying to get as much life out of its tyres as possible and wull not be used on the WCML. So reading between the lines there is opportunity for Tornado to pick up some work with VT, though clearly not the volume RTC would provide.

 

It would seem more likely that VT will promote trips that they run with Tornado themselves rather than run them for RTC.

 

A question to throw in now the A1 Trust will be doing "their own" tours with VT as the operator. Are VT going to be advertising them on their website and be the booking agents, or will the Trust start using their currently redundant '21st Century Steam' brand?

 

I read @acorb as saying that with the two Castles requiring maintenance, an opportunity has come up for Tornado with VT which will give it some work though not as much as if it had stayed with WCR. I could see the 2026 version of the Shap Mountaineer being pretty popular !

 

While I'm super happy about this news, this is a partnership that will work well and I'd certainly be interested in a trip run by Vintage Trains with Tornado but the prices I'm baffled by. £165 for Standard Class is £50 more then normal for Vintage Trains, I can understand the prices increase for a double header as there's two locos in steam and with all the effort required. But £50 more then normal seems unreasonable, however there's probably plenty of reasons behind this so it is what is. Just wonder if this will effect sales.

 

I would hope they would use 21st Century Steam as the booking agent, because VT use Fare Harbor as booking agent which charges a booking fee of just over 5.5%.

Dave B

 

I'm confused too. I can't see anything on VT's booking page for 10th December and only Polar Expresses on Saturdays in December.

Again I'm confused. The only forthcoming double header trip that I'm aware of is this Saturday's Mayflower. How is that relevant to VT trips with Tornado?

 

I’ve missed all this excitement. Despite this being a watched thread I’ve had no notifications for the VT chat above?

 

The 10th is via A1 Steam Trust scroll down for link site https://www.a1steam.com/steamtrains.

The second quoted paragraph relates to difference in cost between VT's own trip and the A1 one, links would have help people follow it.

 

Aha!
On VT's website it's Wednesday 3rd December. I hope VT and the A1 Trust aren't really looking at two different dates.

 

There are going to be two trains to York with different fares. The VT website hasn't caught up yet.