Re: NRC <> Brisbane to Sydney safeworking.

["Richard Wardle" <ricdeb@netmedia.com.au>]

David Proctor wrote in message
<72p4ap$dqi$1@reader1.reader.news.ozemail.net>...
>I agree, Michael - I was just paraphrasing Ebens post.
>
>What p*sses me off is that people say it is unsafe because it uses
>computers, which muck up/does not use signals/relies on drivers to follow
>instructions - when all of this applies to current systems. (I know, there
>are signals at unattended crossing loops - but drivers are often REQUIRED
to
>go through these at stop.
>
>David "The Doctor" Proctor
>daproc.spambait@umpires.com

My biggest problem with train order working David is not the computer but
the variables that the computer cannot handle that require human
intervention like a failure, a last truck breaks away from a 120 long freght
train or the train controller makes an error.
The safest form of safe working is a system that has hidden safe guards like
C.T.C with its track control and the staff.  With a staff system the
starting signal cannot be cleared unless the signaller has the staff with
him.
>
>Michael Walker wrote in message <911215053.332815@woody.hotkey.net.au>...
>>>What if the computer suffers a power out for a fraction of a second and
>>resets.
>>>it then boots up and issues an oerder for a train that is heading for the
>>train
>>>it issued an order to before it went down.
>>>
>>>and computers do make errors. what if it gets a fualty byte in it's
memory
>>and
>>>that corrupts the data been carried?
>>>
>>Without knowing for sure how cheap and nasty the computer systems used for
>>present safeworking systems be they CTC or SAW, a lot of these errors can
>be
>>avoided. If a computer suffers a power out for a fraction of a second, I
>>presume the UPS installed would keep the PC running until either the power
>>resumes or it can be safely shut down.
>>
>>When it is rebooted, I presume the database of logged track orders would
>>come back on line too and take note of where trains are supposed to be and
>>not issue conflicting orders.
>>
>>As for corrupt data, I would assume that if it is on the hard drive that
>>whatever system is set up for mirroring the hard drive be it a redundant
>>drive or redundant server would use the correct data. If a redundant
server
>>is running, that would also fix your memory (and probably your rebooting)
>>problem.
>>
>>I think I can safely assume that a computer system is just as reliable as
a
>>manual system from a fault tolerance point of view. Indeed, I would think
a
>>computer safeworking system is more reliable as it won't permit ANY
>>incorrect moves if set up correctly, unlike a manual system. I know one
>>station that shunts trains without the driver always having the staff,
even
>>though they need to go up and back part of the single line. The reason
>being
>>that when the train gets to the yard, the driver has to walk back to
return
>>the staff. The driver can't be bothered and the station staff get annoyed
>>because they have to wait for the previous train on the single line to get
>>to the end before they can get a staff and then hold up the next down
train
>>whilst they shunt. If the timetable is running behind, it creates all
sorts
>>of problems, then Metrol give them a please explain. Plus it makes the
>>current timetable difficult to keep.
>>
>>So when the driver gets to the yard, he gives a toot and the staff is sunk
>>by station staff (assuming it was removed at all - in this case the driver
>>relies on the fact that the safeworking officer won't release another
staff
>>whilst he is on the single line). Is this a good example of the safety of
a
>>manual safe working system?
>>
>>
>
>