[treis]

I still can't quite grasp why local governments haven't taken the hint and reworked their local bus systems. Uber&Lyft have shown how much better app and on demand service is. But we're still circulating buses and spending billions on heavy rail.

[tfehring]

On-demand service is drastically less efficient than fixed-route service. That's the main reason Uber/Lyft is already several times more expensive than bus or train service for the same route, and it would be even more expensive if Uber/Lyft drivers made the same total comp per hour as bus drivers including benefits, fuel, vehicle depreciation/wear and tear, insurance, etc. Not to mention the additional congestion and pollution you'd get if you replaced every bus during rush hour with 25 cars or whatever.

[treis]

It does not really come out to be that much more. As an example, each unlinked trip costs $5 for my local system. Most trips will take a transfer so a realistic estimate of the per trip cost is $8-10. Looking back at my Lyft history most of my trips cost $8-15 so very comparable.

It looks dramatically different because buses are subsidized while Uber/Lyft is taxed and for profit. To continue using my city as an example, farebox recovery rate is ~30% and that doesn't include capital expenses. The difference is made up by tax revenue. On the other hand out of my fare, ~10% is sales tax, Uber/Lyft take another 20-40%, leaving 50-70% to go to paying the driver for the ride.

The problem ultimately is coverage. If you want to provide frequent service to all parts of the city you end up running a bunch of near empty buses. That kills the efficiency benefits of the buses and when you look at the system as a whole it ends up being about the same as on demand.

[bayindirh]

I don't think buses and rail systems should be phased out in cities. In every city I've seen, buses and rail carry a lot of people in a much more cleaner and efficient manner.

Most, if not all rail is electric. Buses are transitioning to electric in Europe. Geneva has a very nice battery-less (supercap based) bus fleet which carry a lot of people continuously.

No car based transportation can reach the same density.

Taxi has its uses (speed, precise location, etc.), but is no replacement for mass transportation.

[treis]

Few places need the theoretical density that buses provide and even less need the level that rail provides. Most of the time they're running at 10-20% capacity. Or at least that's what happens in my city.

I'm not talking about taxis. Something shuttle bus or van sized. We don't need to put 200 cars worth of people in trains to solve traffic issues. Putting 200 cars of people into 30 shuttle buses is good enough.

[bayindirh]

> Most of the time they're running at 10-20% capacity.

That's not what I've seen in Amsterdam, Geneva, Stockholm, Barcelona, Istanbul and my hometown. They work at least 30% capacity all day long, and when the demand is low, the bus numbers drop in all lines, but you never wait more than ~15 minutes for a new one, in most lines. They also reduce the number of trains on rail systems when the demand is low.

> Something shuttle bus or van sized. We don't need to put 200 cars worth of people in trains to solve traffic issues.

It's not a traffic issue. In modern Europe, it's not about solving the traffic. The governments focus on providing ways people to reach where they want. When you enable it, people doesn't crave for their cars, they use the cheapest and most practical way. In Barcelona, you're almost always walking distance from a underground station so, you don't need a car most of the time.

In my city, due to its geography, it's not possible to have an underground network that dense (at least where I live) and my office is not on a easy to reach place via public transport. If it was the opposite, I'd happily leave my car at home, but I can't.

OTOH, all the public transit lines are working at least 50% capacity all day long, carrying people around the city, and in some places you can literally go from door-to-door using public transit only.

[sdenton4]

We don't plan road building based on off-peak usage numbers; why would we plan for rail that way?

[sjsamson]

> Few places need the theoretical density that buses provide and even less need the level that rail provides.

Almost every major urban city and metropolitan area needs buses and rail service to function effectively, the population density requires it, and it is the proven, mature, and scalable solution to providing high capacity transport. Massive roads used by primarily single occupant vehicles, is the non-scalable, inefficient solution.

> Most of the time they're running at 10-20% capacity. Or at least that's what happens in my city.

Just because you see <i>a bus</i>, at <i>some</i> time of the day being underutilized, is not indicative of much. That same bus on the same day/shift could be at or near capacity, earlier or later in day carrying commuters, or students that got out of class, etc. making it worthwhile.

Capacity planning for a transportation network/system means it has to be sized for its maximum or peak demand (typically during commute hours), so it also makes sense to use its excess capacity during off-peak hours given it is largely paid for to meet peak demand (buses and trains sitting around midday depreciating, and operators being paid a full shift to do nothing, is a poor use of high value assets and resources). This is similar to other systems like energy, telecom, etc.

> I'm not talking about taxis. Something shuttle bus or van sized. We don't need to put 200 cars worth of people in trains to solve traffic issues. Putting 200 cars of people into 30 shuttle buses is good enough.

Same point above about sizing for peak. Additionally this does not match modern fleet management best practices, which would strive to minimize the number of vehicle types in the fleet. Typically this is a standard ~40 foot single level bus. Possibly with additional types (longer articulated or double decker) for lines with higher demand/passenger loads, if needed. A common fleet type minimizes driver and mechanic training, makes buses and personnel more interchangeable and operable across the entire route network, creating efficiencies and economies of scale. These are the same reasons why Southwest Airlines and Ryanair exclusively fly B737s.

Upwards of 80% of the cost structure of providing bus or train service (the marginal cost of running an additional bus/train) is dominated by labor costs in the USA and other developed countries, where labor is expensive and capital, broadly defined, is cheap; so capital replaces labor where possible (developing countries tend to be the opposite). Or to put it another way, 20-40% of the cost structure is the cost of the vehicle. Using smaller, cheaper, low capacity vehicles does not significantly reduce the underlying cost structure, and in fact will increase inefficiencies and costs elsewhere in the system.

An excessive number of small vehicles instead of a reasonable number of large vehicles, where each vehicle requires a driver, will maximize inefficiencies and create diseconomies of scale, resulting in higher overall system costs, lower transport capacity, and higher congestion and pollution in space constrained urban areas. Efficient use of space in constrained urban areas is key attribute of transportation systems [0][1][2].

[0]: https://drive.google.com/file/d/1gl3bVsV3Kcl_RfIFsJ8iZ7dEEEg... [1]: https://en.wikipedia.org/wiki/Passenger_load_factor#/media/F... [2]: https://en.wikipedia.org/wiki/Passenger_load_factor#/media/F...

[treis]

You're arguing theory and not what's happening in the real world. As an example, your cite says that rail has a theoretic density of 60-90 thousand an hour. But in my city we're spending several billion dollars to build a new rail line that is expected to carry 20,000 people a day.

Theory and practice are different. And in most places in the US your theory doesn't translate into practice.