January 2011 in Network Rail and Suppliers
A reporter has contacted CIRAS with numerous concerns about the new Class 378 units operating on the London Overground network.
It appears that many of the issues stem from software problems however it seems management are adapting policies and procedures rather than trying to address the software issues. The concerns are outlined below:
Automatic Warning System (AWS) failures
There has been an increase in the volume of right side and wrong side AWS failures in the London Euston area. Drivers are experiencing failures on a daily basis and many have given up reporting the issue as it seems no action is being taken. The reporter is unsure if the problem is rolling stock or infrastructure related.
Could LOROL and Network Rail clarify the cause of the problem, whether it is rolling stock or infrastructure related and comment on any action being taken?
A software problem where the train's onboard computer thought the train was still moving, when in fact it was stationary at a platform, led to a change in driving policy. Drivers are now instructed to use step three (originally used for stopping at signals) when stopping at platforms. Step three involves Drivers putting the master switch in neutral and pressing the Driver Reminder Appliance (DRA) button. The same is now being done when stopping at a platform, but instead of the DRA button being pressed, Drivers open the doors. However, Drivers have now become so habituated to this practice it is leading to multiple incidences of Drivers opening the doors when stationary at a signal, rather than at a platform.
Do LOROL have any plans to change their driving policy or is this new procedure an interim measure until the software problems are resolved?
Changing power supply
Another software problem has occurred in that the onboard computer thinks the train is not receiving enough current from the conductor rail when stationary. To alleviate this problem when trains terminate at Euston station, Drivers are now instructed to raise the pantograph to connect to the Over Head Line (OHL) so the train receives dual voltage. Once the Driver has changed ends and is ready to depart the station, the pantograph has to be lowered as dual voltage is no longer required. The concern is that the new Class 378 units are not fitted with a dual voltage alarm, which warns Drivers they are connected to both the conductor rail and OHL. Leading Drivers to easily forget their pantograph is up, depart from the station and risk smashing the pantograph on the roof of a tunnel.
Again is this new procedure an interim measure until the software problems are resolved? If no, would it be possible to install a dual voltage alarm to the new units?
Before the new Class 378 units were brought into service, Drivers were given 'driver operated only' training on units made up of three cars. However, when the new units were brought into service they were increased to four cars in length. As a result, there have been multiple incidences of Drivers stopping short of platforms, stopping the train at the three car stopping board instead of the four.
The reporter is aware that reminders have been issued to Drivers about this but it is still ongoing?
A further software problem has occurred meaning Drivers have been instructed to operate their trains in 'ice mode'. The detection of 'spikes' by the onboard computer, means that it is shutting the system down if a particular number are detected over a certain distance. 'Ice mode', used in cold or snowy conditions, allows more 'spikes' to be detected meaning that the onboard computer does not shut down. The reporter is concerned about the use of 'ice mode' when the conditions do not require it.
Could LOROL and Bombardier outline:
Network Rail would like to thank the reporter for bringing their concerns to our attention.
Both Network Rail and LOROL are completely aware of the issue and reported problems that have been occurring in relation to Automatic Warning System (AWS) and the Class 378 train. Both companies are working closely together to identify the solutions that will successfully resolve the current issue.
The problem is complex as the trackside AWS on the DC lines approaching Euston is required to work with both AC and DC trains which have unique AWS receivers that require different signal threshold levels to operate successfully.
Network Rail and LOROL Engineers have identified a number of
train borne and trackside measures, and is considered by the
relevant technical specialists in both companies, which will
address the current situation. These will be implemented over the
coming weeks, following which their effectiveness will be
Both Network Rail and LOROL are committed to resolving the issue and both organisations wish to confirm that Drivers experiencing AWS faults should continue to report them. The reporting of any AWS fault is important to establish that the solution is effective and addresses the AWS failures currently being seen by Drivers.