3 ms·

I care about privacy quite a bit but there doesn't seem to be much controversy here. Yes it's creepy that O2 has pretty tight demographic data and shares some

by uoaei 3y ago

I care about privacy quite a bit but there doesn't seem to be much controversy here. Yes it's creepy that O2 has pretty tight demographic data and shares some of that information with TfL, and we have to assume they're doing that for other more nefarious customers with fewer privacy restrictions. But telecoms being involved in that kind of shady business is old news. I'm all for improving the information (given it's private) that public transit service lines and scheduling planners have access to if it means smoother transit and especially less dependence on cars.

vidarh 3y ago
> I'm all for improving the information (given it's private) that public transit service lines and scheduling planners have access to if it means smoother transit and especially less dependence on cars. My dream is a CityMapper style app used to offer "virtual public transit" as the means to both collect this kind of info and actively use it to make transit better. That is, say TFL were to offer up "virtual bus routes" that may or may not coincide with the real ones, and that feed other transit hubs. Where they coincide with regular routes, the selling point would be that they'd hire cars to handle excess demand or gaps in service so you can guarantee pickup from any bus stop within x minutes as long as you plan a journey in the app. What's great about the tube is knowing you can just show up and there's "always" a train soon. Getting something like that for busses would be fantastic, and you can offset some of the cost of guaranteeing more frequent service by actually cutting the frequency of full-size scheduled buses on less used stretches as long as you guarantee on average shorter waits when people need it. Maybe with some limitations - e.g. if not a "real" bus you might be shunted onto regular services at the nearest transit hub. Where they don't coincide with regular routes, you use first/last mile data from the planning to try to guess at new routes to offer, and so when people book, they might get offered to go to a nearer pickup point that isn't a bus stop but where you're guaranteed to be picked up. There are some services sort-of like this, but usually as a separate thing, and it'd be great if it's fully integrated as a means to plug holes in service and test route changes. I think there's a lot of chicken and egg going on with transit where likely viable routes aren't available because testing them out is difficult and expensive and gathering data on which routes might be useful likewise. If you can fill them with smaller, cheaper cars and dynamically test, I think it'd have the potential to get a lot more people to try to stick with public transport. Especially if it also offer "fallback" or combined options involving booking a taxi for all or part of it when even the virtual routes can cut it. CityMapper is great for the planning in London, but that ability to offer "virtual routes" at scale is what would really change things.
- bobthepanda 3y ago
  that would be good for direct routes, but public transport is a bit different in that you can offer much better connections with a pair of frequent routes and a transfer, especially if the transfer is no-cost as it is with the bus hopper fare in London. Consequently new routes have a cascading effect on a network and are treated gingerly, particularly if the budget remains fixed and you are robbing Peter to pay Paul. for the most part mapping services like Google Maps often provide the public transport times, bike share, ride share, etc. within the same page and query where I live.
  - vidarh 3y ago
    My thinking is that it's particularly for routes where it would still involve changes that there is most to gain. You're more likely to be willing to put up with public transport if there's a quick way to get to a hub. An example near me is that the main bus routes follow one of 3 two main roads between two major stations near me. If you're in the middle or want to cross between those two, it's usually faster to walk. Especially as the buses are often full during rush hour. If you were able to both offload a portion and offer a more attractive alternative for those living in the middle, even just between those two stations, it'd plug a hole. That route would break even if you managed to put on average 4 passengers per minibus paying the same as the bus fare based on Uber and minicab prices. Given bus routes are subsidised, that's a pretty decent starting point even if you drive increased demand to other parts of the network. On the other end of the scale, there are places a bit further out to me where the buses run every half an hour, if you're lucky, and winds it's way across a large area to cover a much as possible with as few buses as possible where the demand is at the same time so low that you could meet it far better with near-on-demand minibuses that'd skip. I think you could do a lot here without actually increasing budgets, by scaling down services that are currently kept going because they're seen as necessary, but replacing them with something more dynamic. > for the most part mapping services like Google Maps often provide the public transport times, bike share, ride share, etc. within the same page and query where I live. Google Maps is very sub-optimal in London compared to the Citymapper model which is to hand-hold you through the entire journey (tell you which platform to go to, vibrate when you're close to a stop etc.), but that's not really the point. The point would be to be able to plan your journey in a way that ensured the data could feed straight into bookings as well as into end to end demand forecasting of location pairs instead of just which journeys people take when they don't have any other alternative.

by uoaei 3y ago

vidarh 3y ago

> I'm all for improving the information (given it's private) that public transit service lines and scheduling planners have access to if it means smoother transit and especially less dependence on cars. My dream is a CityMapper style app used to offer "virtual public transit" as the means to both collect this kind of info and actively use it to make transit better. That is, say TFL were to offer up "virtual bus routes" that may or may not coincide with the real ones, and that feed other transit hubs. Where they coincide with regular routes, the selling point would be that they'd hire cars to handle excess demand or gaps in service so you can guarantee pickup from any bus stop within x minutes as long as you plan a journey in the app. What's great about the tube is knowing you can just show up and there's "always" a train soon. Getting something like that for busses would be fantastic, and you can offset some of the cost of guaranteeing more frequent service by actually cutting the frequency of full-size scheduled buses on less used stretches as long as you guarantee on average shorter waits when people need it. Maybe with some limitations - e.g. if not a "real" bus you might be shunted onto regular services at the nearest transit hub. Where they don't coincide with regular routes, you use first/last mile data from the planning to try to guess at new routes to offer, and so when people book, they might get offered to go to a nearer pickup point that isn't a bus stop but where you're guaranteed to be picked up. There are some services sort-of like this, but usually as a separate thing, and it'd be great if it's fully integrated as a means to plug holes in service and test route changes. I think there's a lot of chicken and egg going on with transit where likely viable routes aren't available because testing them out is difficult and expensive and gathering data on which routes might be useful likewise. If you can fill them with smaller, cheaper cars and dynamically test, I think it'd have the potential to get a lot more people to try to stick with public transport. Especially if it also offer "fallback" or combined options involving booking a taxi for all or part of it when even the virtual routes can cut it. CityMapper is great for the planning in London, but that ability to offer "virtual routes" at scale is what would really change things.

bobthepanda 3y ago
that would be good for direct routes, but public transport is a bit different in that you can offer much better connections with a pair of frequent routes and a transfer, especially if the transfer is no-cost as it is with the bus hopper fare in London. Consequently new routes have a cascading effect on a network and are treated gingerly, particularly if the budget remains fixed and you are robbing Peter to pay Paul. for the most part mapping services like Google Maps often provide the public transport times, bike share, ride share, etc. within the same page and query where I live.
- vidarh 3y ago
  My thinking is that it's particularly for routes where it would still involve changes that there is most to gain. You're more likely to be willing to put up with public transport if there's a quick way to get to a hub. An example near me is that the main bus routes follow one of 3 two main roads between two major stations near me. If you're in the middle or want to cross between those two, it's usually faster to walk. Especially as the buses are often full during rush hour. If you were able to both offload a portion and offer a more attractive alternative for those living in the middle, even just between those two stations, it'd plug a hole. That route would break even if you managed to put on average 4 passengers per minibus paying the same as the bus fare based on Uber and minicab prices. Given bus routes are subsidised, that's a pretty decent starting point even if you drive increased demand to other parts of the network. On the other end of the scale, there are places a bit further out to me where the buses run every half an hour, if you're lucky, and winds it's way across a large area to cover a much as possible with as few buses as possible where the demand is at the same time so low that you could meet it far better with near-on-demand minibuses that'd skip. I think you could do a lot here without actually increasing budgets, by scaling down services that are currently kept going because they're seen as necessary, but replacing them with something more dynamic. > for the most part mapping services like Google Maps often provide the public transport times, bike share, ride share, etc. within the same page and query where I live. Google Maps is very sub-optimal in London compared to the Citymapper model which is to hand-hold you through the entire journey (tell you which platform to go to, vibrate when you're close to a stop etc.), but that's not really the point. The point would be to be able to plan your journey in a way that ensured the data could feed straight into bookings as well as into end to end demand forecasting of location pairs instead of just which journeys people take when they don't have any other alternative.