23-12-2017, 05:44 PM
If one goes back to the origins of overlaps when there was no train protection system then the length was to allow for misjudgement, driver microsleeps (although we did not call them that then), etc. History had led us in UK to a long distance being needed for a semaphore home signal (because at night it and the distant were lit by sleepy glowworms!) but for colourlight signals with cautionary aspects then a shorter distance could be used.
Where that distance was "inconvenient" then we allowed the distance to be reduced if the approach speed was low (on the grounds that the driver had more time to react if they misjudged, etc. plus the consequence was probably lower) or if we applied aspect controls on the previous signal to "warn the driver" and also make sure the train was going slower.
When one introduces a system such as AWS then there is an extra prompt to the driver (but we now know they can acknowledge it subconsciously).
Adding a train protection system such as TPWS changes the picture again - and we (NR) have for some time been requiring that the TPWS is effective to stop most trains using the route within the SOD (which may be longer than the overlap). Effectiveness is tested at the worst possible approach speeds so for most train approaches the speed will be lower and the train, if tripped, will stop earlier (and indeed most stop before the signal).
What that allows us to think about is removing the concept of a full overlap, a reduced overlap and a restricted overlap and designing a combined TPWS and overlap where the two combined provide the required level of protection.
Looking afresh at safety challenges and all the measures we have put in place over the years can lead to some surprising conclusions.
Where that distance was "inconvenient" then we allowed the distance to be reduced if the approach speed was low (on the grounds that the driver had more time to react if they misjudged, etc. plus the consequence was probably lower) or if we applied aspect controls on the previous signal to "warn the driver" and also make sure the train was going slower.
When one introduces a system such as AWS then there is an extra prompt to the driver (but we now know they can acknowledge it subconsciously).
Adding a train protection system such as TPWS changes the picture again - and we (NR) have for some time been requiring that the TPWS is effective to stop most trains using the route within the SOD (which may be longer than the overlap). Effectiveness is tested at the worst possible approach speeds so for most train approaches the speed will be lower and the train, if tripped, will stop earlier (and indeed most stop before the signal).
What that allows us to think about is removing the concept of a full overlap, a reduced overlap and a restricted overlap and designing a combined TPWS and overlap where the two combined provide the required level of protection.
Looking afresh at safety challenges and all the measures we have put in place over the years can lead to some surprising conclusions.