Managing dwell times through seat selection when booking

[JKF]

I’ve no idea if this is done already, but could seat selection in booking systems for long distance trains aim to distribute passengers for specific destinations across the length of the train, so that on arrival those leaving or boarding via each door is minimised? This could potentially help reduce dwell times and conflicts within gangways and vestibules. Marginal gains, but would cost very little to implement.

As an example, say there is a six-coach train stopping at Station X. Thirty passengers have journeys booked to Station X, so the booking system aims to put five passengers in each carriage. As each booking is made it picks the next carriage each time, as best as possible keeping the numbers spread evenly across the set. That way around five people leave via each door, rather than ten, twelve or so on from one door and none from another.

Obviously people can choose their own seat, but not many choose a different carriage, and if they do, the next booking is suggested in the same place to try and keep it all balanced.

Walk-up fares won’t be managed, and some large groups travelling together will all want to be in the same coach, but I think it would still have some useful impact.

Other considerations could be that if there’s a particular coach that tends to fill up more by walk-on passengers as it’s by an entrance/exit at a particular station then focus on keeping booked passengers for that destination elsewhere, to balance this out. Also perhaps prioritise keeping passengers for intermediate stations (where dwellings times are more critical) in seats nearer doors so they can depart quicker.

This could be supported by monitoring/verification to see how many passengers use each door at each station, with refinements made to optimise balancing passengers out.

 

[The Planner]

I’ve no idea if this is done already, but could seat selection in booking systems for long distance trains aim to distribute passengers for specific destinations across the length of the train, so that on arrival those leaving or boarding via each door is minimised? This could potentially help reduce dwell times and conflicts within gangways and vestibules. Marginal gains, but would cost very little to implement.

As an example, say there is a six-coach train stopping at Station X. Thirty passengers have journeys booked to Station X, so the booking system aims to put five passengers in each carriage. As each booking is made it picks the next carriage each time, as best as possible keeping the numbers spread evenly across the set. That way around five people leave via each door, rather than ten, twelve or so on from one door and none from another.

Obviously people can choose their own seat, but not many choose a different carriage, and if they do, the next booking is suggested in the same place to try and keep it all balanced.

Walk-up fares won’t be managed, and some large groups travelling together will all want to be in the same coach, but I think it would still have some useful impact.

Other considerations could be that if there’s a particular coach that tends to fill up more by walk-on passengers as it’s by an entrance/exit at a particular station then focus on keeping booked passengers for that destination elsewhere, to balance this out. Also perhaps prioritise keeping passengers for intermediate stations (where dwellings times are more critical) in seats nearer doors so they can depart quicker.

This could be supported by monitoring/verification to see how many passengers use each door at each station, with refinements made to optimise balancing passengers out.

People are just going to board the train where they see fit and walk down the train. I don't see how this could work.

 

[JKF]

People are just going to board the train where they see fit and walk down the train. I don't see how this could work.

Some might, but some will go to the right door, and most coming off will use the nearest exit so perhaps some time gain from disembarking. Every little helps. Setting up a system like this would cost a bit in terms of programming and survey/monitoring, but running cost would be minimal once up and running.

 

[zwk500]

Some might, but some will go to the right door, and most coming off will use the nearest exit so perhaps some time gain from disembarking. Every little helps. Setting up a system like this would cost a bit in terms of programming and survey/monitoring, but running cost would be minimal once up and running.

In theory yes, but I suspect it'll be incredibly difficult to realise the benefits because of passenger behaviour nearly always taking what's easiest for the individual not what's most efficient for the system. Personally I'd rather focus my attention on rolling out the numbered zones nationally with clear, consistent displays of where trains will stop.

Mind you, even just bringing back the 'let people off before you board' mantra would help. Becoming quite a regular occurrence now to arrive at a station with a solid semicircle of people in front of you obviously expecting you to pull off a double somersault with your luggage to get over them. And even when they do stand aside it's only ever wide enough for 1 person, despite there being a very obvious width guide in the shape of the doors.

 

[SynthD]

Other considerations could be that if there’s a particular coach that tends to fill up more by walk-on passengers as it’s by an entrance/exit at a particular station then focus on keeping booked passengers for that destination elsewhere, to balance this out.

Some of the trains I rush onto at the last minute are made of two sets, I’d be mighty disappointed to not have my seat until I change sets at the first stop. Sometimes it’s not even the train nearest the buffers, so I couldn’t reliably count carriages, or have them easily signposted.

Balancing out could be achieved other ways, such as adjusting stopping positions on long platforms. The trains always stop beyond the way out at my home station in one direction.

 

[LittleRedTrain]

In my experience, if you really want to reduce dwell times, the best option is to remove reservations entirely and just let people park themselves in the first free seat they see, rather than wandering slowly down the aisles, trying to find their booked seat and potentially evicting someone already in it, while a queue of frustrated passengers builds up behind them. (Just compare how long it takes to board a busy 4 car CrossCountry train with reservations vs a 4 car Chiltern train without!).

Also, making sure there's adequate luggage space, so there's always a free spot for for someone to quickly stow their case rather than having to do a re-shuffle of the rack to fit it in, again while other passengers are stuck behind them.

Of course, the best way to reduce dwell times is to make sure that trains have enough capacity that you don't have to fight your way through standing passengers in the aisles or vestibules in order to get on or off. (I'm looking at you again XC!)

 

[djox]

n my experience, if you really want to reduce dwell times, the best option is to remove reservations entirely and just let people park themselves in the first free seat they see, rather than wandering slowly down the aisles, trying to find their booked seat and potentially evicting someone already in it, while a queue of frustrated passengers builds up behind them. (Just compare how long it takes to board a busy 4 car CrossCountry train with reservations vs a 4 car Chiltern train without!).

If only CrossCountry would remove reservations until they can actually provide enough seats. We can only dream.

 

[The Planner]

I still don't see why it will make amy difference. People just get on and walk down the train to find their seats. There have been countless times where I have been on a train that has departed and people are shifting people out of reserved seats 5 or 10 minutes later. If you want to reduce dwells, focus on door cycles of units when the train is at a stand.

 

[zwk500]

If only CrossCountry would remove reservations until they can actually provide enough seats. We can only dream.

If anything, reservations are more important when there's no guarantee of a first-come-first-served seat than when they do have enough capacity.

What needs to be better is a clearly communicated policy around swapping seats or occupying no-shows.

 

[Magdalia]

I detect a lack of clarity of thinking in this discussion.

First, dwell times matter far more at intermediate stations than at start and end of journey.

Second, people getting off the train matter far more for dwell time than people getting on. This is because the people getting on usually can't do that until after the people getting off are out of the way. Furthermore, people getting off usually use the nearest door, whereas people getting on can use any door and find their seat after the train has started to move.

So where seat reservations could usefully help to manage dwell times is by managing the people getting off at intermediate stations. If these are distributed evenly along the train then that will help to reduce dwell times.

Take a Kings Cross-Edinburgh train calling at York as an example. If the Kings Cross-York passengers are evenly distributed through the train then they will get off more quickly and allow boarding of passengers joining at York to start sooner.

 

[zwk500]

In my experience, if you really want to reduce dwell times, the best option is to remove reservations entirely and just let people park themselves in the first free seat they see, rather than wandering slowly down the aisles, trying to find their booked seat and potentially evicting someone already in it, while a queue of frustrated passengers builds up behind them. (Just compare how long it takes to board a busy 4 car CrossCountry train with reservations vs a 4 car Chiltern train without!).

I'm not sure this would be faster because people tend to fill up the seats nearest the doors first so it'd take ages for people to get off the platform into the train itself. Also on longer-distance trains you'd still have all the luggage kerfuffle and people hovering while they decided if a seat is actually free or there's an occupant who's not clearly visible.

 

[JKF]

I still don't see why it will make amy difference. People just get on and walk down the train to find their seats. There have been countless times where I have been on a train that has departed and people are shifting people out of reserved seats 5 or 10 minutes later. If you want to reduce dwells, focus on door cycles of units when the train is at a stand.

Will it not make a difference to the time taken for disembarking if booked passengers for a destination are spread evenly through the train? And I can’t be the only person who will walk down the train to find my carriage to avoid having to squeeze through several crowded coaches.

 

[zwk500]

I detect a lack of clarity of thinking in this discussion.

First, dwell times matter far more at intermediate stations than at start and end of journey.

Second, people getting off the train matter far more for dwell time than people getting on. This is because the people getting on usually can't do that until after the people getting off are out of the way. Furthermore, people getting off usually use the nearest door, whereas people getting on can use any door and find their seat after the train has started to move.

So where seat reservations could usefully help to manage dwell times is by managing the people getting off at intermediate stations. If these are distributed evenly along the train then that will help to reduce dwell times.

Take a Kings Cross-Edinburgh train calling at York as an example. If the Kings Cross-York passengers are evenly distributed through the train then they will get off more quickly and allow boarding of passengers joining at York to start sooner.

Part of this is room in the train, at busy stations (York is a classic example) the people queuing for the door will be stopping people getting their luggage out, so they have to wait for the first few people to get off before they can grab their bags, slowing down exit. Part of it is in the boarding distribution, because people are trying to stow all their luggage then find their seat without thinking to let people actually get off the platform first.
However I'd contend the worst impact on dwell times comes from people waiting to board surrounding the doors before the train's even stopped such that a person alighting has nowhere to go without barging into people. It can take several seconds, including a 'Excuse me' or hand gesture from the alighting passengers, before anybody will stand to one side and even then they leave a tiny sliver of room rather than a nice wide corridor. Next after that is people alighting who put their bags down and immediately spend a couple of seconds fiddling to get the handle up rather than getting clear of the doors so that others can get off behind them while they sort their luggage out.

 

[Magdalia]

Part of this is room in the train

What matters here is good design. The exemplar for a train designed for managing dwell time is the class 700. This doesn't just have to deal with lots of commuters it also has lots of passengers with luggage going to and from Gatwick and Luton airports.

Most modern long distance trains are very badly designed for getting out and in at intermediate stations. I blame this on the obsession with making the train more like a plane. Most plane journeys do not have intermediate calls. In the no blame culture it is all the fault of the man with the beard.

A modern example of a long distance train breaking away from the plane model is the class 745.

However I'd contend the worst impact on dwell times comes from people waiting to board surrounding the doors before the train's even stopped such that a person alighting has nowhere to go without barging into people.

Better signage helps here, and again Thameslink leads the way. The core stations have platform markings for where each set of doors will be, and footprints to show people where to stand so that there is a gap for the people getting off.

 

[The Planner]

Will it not make a difference to the time taken for disembarking if booked passengers for a destination are spread evenly through the train? And I can’t be the only person who will walk down the train to find my carriage to avoid having to squeeze through several crowded coaches.

If you could force people to do it, then it would reduce the chances of over dwelling, but people don't work like that.

 

[Snow1964]

Yes, it would be possible to change algorithms to include a disembarking passengers factor at intermediate stations when system chooses seats, to try and even it out, so not getting loads alighting off one coach and few next coach.

Down here in South West all platforms are marked with zones, but I have never seen them listed on the itinerary next to seat and coach number to help passengers be by correct coach ready for boarding, and avoid passengers having to walk down (and of course some will walk in opposite direction because their seat is other way from where they boarded which causes busy aisles and congestion), those with big cases having to fight their way along to correct coach is problem.

Taking my area as an example, although you can buy advances they all say sit anywhere, no specific seat (even on trains taking 3+ hours) so there is lot of inconsistency nationally.

But overall yes I agree, could be lot more logic applied to reservation systems to make passenger churn work better and minimise dwell times.

 

[zwk500]

What matters here is good design. The exemplar for a train designed for managing dwell time is the class 700. This doesn't just have to deal with lots of commuters it also has lots of passengers with luggage going to and from Gatwick and Luton airports.

Indeed, and station design is an imporant factor here as well. Gatwick Airport station has some quite narrow sections along the platform buildings, which counteracts the good work of the 700s in having wide entryways.

Most modern long distance trains are very badly designed for getting out and in at intermediate stations. I blame this on the obsession with making the train more like a plane. Most plane journeys do not have intermediate calls. In the no blame culture it is all the fault of the man with the beard.

?

Better signage helps here, and again Thameslink leads the way. The core stations have platform markings for where each set of doors will be, and footprints to show people where to stand so that there is a gap for the people getting off.

Interestingly, my most recent experience of this was at St Pancras Thameslink, where due to the position of the markings where the doors would be, passengers looking to board decided the best thing to do was to swarm in front of it and not stand aside. For reference, it was an Easter Weekend journey so very much leisure passengers only, including a fair bit of luggage. I suspect Thameslink commuters are a bit better, but I also regularly experienced this on weekday morning commutes at Milton Keynes, especially so post-COVID restrictions.

 

[djox]

I would welcome non reservation carriages and reservation carriages, which are clearly signed on the platform screens, that way if you have no reservation you can get a carriage and know you can sit anywhere and be able to board quickly, it should also help as the carriages which are for reservations won't have people trying to sit in your seat so should speed that carriage boarding up too.

 

[Haywain]

The exemplar for a train designed for managing dwell time is the class 700.

That may be so but you wouldn’t want to be on one of those for a long distance journey.

 

[NCT]

Are alighting passengers ever bunched up in particular carriages? This seems like a deterministic solution looking for a probabilistic problem.

Reservations are awful for slowing down boarding. German ICEs are notoriously bad for this. Passengers squint for reservation information and turf passengers out of seat 03 (sometimes erroneously) while the 10th boarding passenger is stuck on the platform unable to board.

A more fundamental aspect is that only true intercity services should have end-door intercity stock, and intermediate stops with standard 2-minute dwells should not be serving markets bigger than 10-on-10-off per door. This is one of the aspects British intercity gets right - large markets are served with originating / terminating trains so passengers can board and alight at their leisure, or you have extended dwells at the large cities (e.g. Bristol extensions to Weston, Leeds extensions to Harrogate, Edinburgh extensions to Aberdeen).

Things like the TransPennine Express whose purpose is really to carry Huddersfield - Leeds commuters should use (high-spec) commuter stock with 1/3 2/3 big doors.