When conducting training courses for new schedulers, a specific part of a network I once worked on is often used to help explain a concept or two. This is because of the amount of lessons I personally learnt early in my career as a result of operating this one simple route as part of a greater network.
Firstly I should help you by explaining the scenario as I first encountered it. As part of a city network there was one estate, sandwiched between major trunk roads, being served by 2 circular routes. The service 5 & 5a each operated with a 30mins headway that combined to give a 15mins headway across their common portion of route.
These services each had 2 vehicles allocated to them all through the day, however, with a 45mins round journey time, it should have been possible (on paper at least) to share resources and operate these using just 3 vehicles. This had been the case in the past, but due to persistent late running a 4th vehicle was added to improve reliability.
Due to a fall in revenue, it was no longer deemed sustainable to continue using 4 vehicles on these services. However, due to the problem of late running we knew operating a 15mins combined headway with 3 vehicles was not going to work either. So the easy answer, surely, would be to drop it to a 20mins combined headway.
Now this wasn’t really as straight forward as that, as a 20mins combined headway is a reduction in service. This would also lead to a 40mins headway on each route, meaning in some hours, 2 journeys operate on the 5 and in others only 1; and the opposite for the 5a. Therefore, the next step in the process would be to drop one service to make sure every journey operates the same way around the estate to avoid any confusion for potential customers.
Again, there could be a simple answer here; just drop the 5a. All the other services on the network are the 1, 2, 3 etc. without any ‘a’ or ‘b’ variants, so keeping the 5 seems the most logical choice, right?
Well, instead of just settling for the simple answer, it was decided that we should investigate the performance of each service to make sure that the right route was being taken out.
When we studied the performance of each service it was found that the service 5 in most cases returned to the City centre around 5mins later than scheduled, whereas the 5a was always within a couple of minutes of its planned arrival time. This therefore suggested that the 5a was the more reliable of the 2 services and should be the route every journey should take going forward.
Interestingly, the reason for the service 5 always being late is so straight forward it was never considered an issue. Both the service 5 & 5a had the same round journey time of 45mins as they covered the same route, albeit in opposite directions. However, that was the problem. Due to the service 5 operating a CLOCKWISE loop, it had to perform mainly right turns at junctions and the more reliable 5a had mainly left turns. Unfortunately for the service 5, these right turns were being attempted on main trunk roads, across heavy oncoming traffic. Despite there being traffic lights on 4 of these turns, none of them gave any priority for vehicles turning right. Some did however allow priority for left turns which was of great benefit to the service 5a.
After looking at the performance of each route and deciding to make every journey a 5a, we could now also have the confidence that operating a 15mins headway with only 3 vehicles was possible. This was due to knowing this service had the correct round journey time of 45mins and was reliable with it. For this reason we increased the headway back up to every 15mins, instead of the every 20mins service we had planned, as late running shouldn’t be an issue.
So this was only a small and simple change, but there are certainly lessons here to be learnt;
1) Know your networks. The issues regarding right turns and late running was common knowledge amongst the drivers, however, as the extra resource was already there to remedy this, no journeys were lost and therefore this hadn’t been considered higher up. Being as knowledgeable as the drivers would have been a great benefit here.
2) The easy answer isn’t always the best answer. As was seen here, operating at a 20mins headway initially seemed the easiest and best way forward. However, this forced us to look at operating only 1 route, to avoid the unfriendly 40mins headway. This forced change meant a better 15mins headway could be achieved instead of the ‘easy’ 20mins answer, by ensuring that the right route was operated.
3) Learn from your mistakes. Although it had happened before my time there, the addition of the 4th vehicle onto these routes had been the wrong decision. It was the easy answer at the time (see lesson 2!) but all it achieved in the long run was an increase in cost. The key problem persisted, although it was now masked by the inefficient added time the extra resource brought with it.
Here we have 3 important lessons gifted to us thanks to one simple network review. But this service has more lessons to offer, like this one here.
Keep a look out for more lesson from service 5 in the future.
Kieran Proctor 