Hacker News Reader

Comment by HonestOp001 | Hacker News Reader

It is the future, getting the large inefficient buses off the system is the goal.

His thesis that the vehicles meander and no technology will fix that is refutable in that technology can identify opportunities to minimize inefficiencies. A good algorithm will create efficient routes for the flexible transit.

Finally, the author is a public transit consultant. He is not going to be good for a reference when he is on one side of an item. He may be a subject matter expert, but he is blinded to what can be accomplished. Judging from other people’s comments from other countries, my thesis is more accurate.

View on HN

Flexibility and efficiency are often at odds.

Fixed route service is FAR more efficient than any sort of micro transit service can ever be, unless we are looking at areas with very low demand.

The most efficient algorithm is always going to be having people gather in one spot to be picked up and dropped off rather than driving all over to different points.

>Fixed route service is FAR more efficient than any sort of micro transit service can ever be

Theoretically fixed route service is far more efficient. Practically, it's not because it takes a lot of demand to run at high frequency and high utilization. As an example, I live in Atlanta and it costs the system $9 per bus ride these days. And given the nature of the system and the city it usually takes people more than one bus ride to get where they need to go.

Lyft/Uber usually comes out to be slightly to much cheaper excepting long cross-metro rides. The problem is that I don't get my $15.50 subsidy.

You aren’t making an apples to apples comparison though, your assertion is that it costs the system $18 for you to take a trip, but you’re only looking at the fare charges by Uber/Lyft. Uber and Lyft continue to operate at a loss, and are therefore subsidizing their rides as well. Additionally, they are able to maintain a contingent workforce that a public transit agency would likely never be able to get away with. Uber and Lyft are directly subsidized by investor money, in addition to the lower wages earned by drivers who also cover the capital expenditure for vehicles and cover the maintenance costs.

And final point, mass transit demand still hasn’t returned to pre pandemic levels. Obviously, it may be the case that we have seen a long term shift in demand, but the fact remains that the current system was built out/provisioned for a higher level of demand, so it’s not surprising that the cost structure is higher right now than it should be. Pre pandemic it looks like bus/rail cost Marta around $4.50 per trip.

>You aren’t making an apples to apples comparison though, your assertion is that it costs the system $18 for you to take a trip, but you’re only looking at the fare charges by Uber/Lyft.

The $18 is operating cost and doesn't include capital costs. Lyft is EBIDTA profitable and Uber likely is on rides too. Their fares are enough to cover the operating cost of the system. Comparison is as apples to apples as it can be.

>Pre pandemic it looks like bus/rail cost Marta around $4.50 per trip.

Even still. To use a practical example, my kid's day care is 4 miles away. I would need 2 buses to get there by MARTA so $9 if we assume pre-pandemic costs. Lyft will take me there for between $8.93 and $14.17 with various options on shared + wait time. For a solo car for a pick up in 4 minutes it's $11.01.

Lyft will take me 9 minutes to get there while MARTA will take 46 minutes if I time everything perfectly. More realistically it's an hour which is barely faster than walking the 4 miles.

Lyft’s adjusted EBITA was $55 million in Q1, given their active rider number stands at ~19 million, at best they are likely offering rides at cost.

That still leaves out the costs incurred by the drivers. Opex for transit agencies includes maintenance, insurance and driver salaries and benefits which are much more costly than it is for Lyft.

You are right that sprawling metro areas are very well suited for mass transit, though. There does need to be a sufficient amount of demand in order for it to work.

The fact of the matter is that the main problem is a land use pattern that requires over $9 to travel to and from a child’s daycare.

Are buses that inefficient? Why is the objective to remove them, and by what measure are they inefficient?

Buses require expensive drivers and so they are the most expensive form of transit. A small bus costs almost as much as a big on to run. A train costs more per trip, but a train typically has many more people on. In theory you could run a train for service where a small bus would work, but since nobody does that the fact that a train would cost a lot more for that service is irrelevant.

Buses are far from the most expensive form of transit. High-speed ferryboats or helicopters probably claim that prize. I'm pretty sure airplanes are also more expensive, as are subways if you include the cost to build them in the first place, as well as cars if you include total cost of ownership.

Cost is typically measured per passenger km or some such. Ferryboats can come out cheap if there are enough people on them. I'm not aware of anyone using helicopters for mass transit. Airplanes are very cheap for long distance service.

> A good algorithm will create efficient routes for the flexible transit.

You assume there is such an algorithm. That assertion so far has proven false, and I see no reason to believe that any algorithm can overcome the fundamental problems.

Reply on news.ycombinator.com