6 ms·

> “Mostly this would be for freight,” says Mark Patton, who oversees transport plans for the Columbus region. Paul Judge, who runs a thriving plastics manufact

> “Mostly this would be for freight,” says Mark Patton, who oversees transport plans for the Columbus region. Paul Judge, who runs a thriving plastics manufacturer, Axium, in the city, says he would welcome it if it meant he could run a big factory, cheaply, in one spot and ship his billions of units of shampoo bottles, hand-sanitisers and the like to markets in the region. This makes zero sense to me. Surely, there must be a law of decreasing returns in supply chain latency. How could reducing delivery times times to hyperloop levels possibly provide a reasonable pay-off period, given the considerable cost of building and maintaining a subterranean vacuum tube train network?

richardw 6y ago
Let’s think it out (as in: add your thoughts). If it’s for goods, you don’t need a human size pod. You can scale pods to a size that represents goods that make sense to transport quickly. Obviously you want benefits that come from standardisation, like containers give you. Fully automated transport between all major cities for anything less than a metre diameter? I think that looks like the future. Mail, Amazon, fresh produce, run as a stream rather than in large batches, automatically routed, staying off highways. More efficient stock management that can react to surges in demand. Charge for it with spot pricing so high value items get priority. Edit: produce sounds a bit silly but taken to extremes might be interesting. When you deliver a ton of lettuce, it’s sold as a stream. Someone gets the fresh stuff and someone gets the three-day old stuff. If you can reorder every 30m, everyone gets fresh. Smaller shelves and stores, maybe no store except online.
- ashtonkem 6y ago
  There are two huge problems here: First, very little freight is latency sensitive. The vast majority of cargo is cost sensitive, which is why most bulk goods coming from abroad arrive via cargo ship rather than airplane. There are a ton of companies that’ll accept days or weeks of delay to reduce overall cost; so a loop must compete with rail freight and semi trucks on cost. Theoretically a cargo hyper loop network could compete with domestic air freight, but I find the idea that air freight alone could cover the cost of a hyper loop network hard to buy. This is doubly true when you consider the fact that air freight shares costs with commercial passenger travel, both in terms of infrastructure and vehicle development. Second, they’re already working on automating existing railroads. There’s no particular reason why a hyper loop is necessary to automate a train, and we’re already seeing some of the very long haul train routes in Australia being automated. This comes with the benefit of much lower development time and cost, since it allows for the reuse of the most expensive bit of infrastructure: the rails themselves.
  - richardw 6y ago
    I never mentioned bulk goods. I mentioned items that are time sensitive. Most e-commerce is latency sensitive. I recently started getting groceries on an hour delivery. It’s a game changer. Amazon could probably power a network like that by itself. What would a 2 hour Amazon delivery to most cities do? What markets does it enable access to? How do you compete with that? It’s too easy to think of why something won’t work. Rather think “is there a customer and a product for which latency is important but is unfulfilled by current infrastructure”. Don’t use my examples, make your own. Saying “no” to that question is to bet against human desires and ingenuity. Freshest (anything) from 200km away within 2 hours. Will top restaurants pay for that? Latest widget from a warehouse 500km away in 2 hours. Allows Amazon to compete with local stores. Allows medical supplies to get there in a surge. I’m sure you can think of something.
    - ashtonkem 6y ago
      > Don’t use my examples, make your own. That’s ... not how arguments work. You’ve got to provide your own examples. I’m not going to do your work for you. > Saying “no” to that question is to bet against human desires and ingenuity This is genuinely a bad way to approach these things, as you’re presuming that it’ll work based on enthusiasm and not evidence. This exact same logic would’ve led you to advocate for flying cars, as you would’ve assumed that ingenuity would triumph over the practical issues that have so far prevented them from working. Ingenuity is great, but market demand actually pays for the bills. And predicting exactly what the market will do is impossible, but given the fact that this will cost trillions of dollars to actually build, it should be possible to predict use cases that’ll bring in billions of revenue. > Allow Amazon to compete with local stores. Amazon is already crushing local stores on price. It turns out that consumers value delivery speed but not as much as they value low cost. A hyper loop might improve the speed of goods between Amazon warehouses (or run fewer), but it would do nothing to improve the “last mile” problem, which is where all the cost is. Drone delivery + a hyper loop would actually be faster; but I’m dubious that this use case alone covers said trillions of dollars of cost; especially when drone delivery from local warehouses would work too. > Will top restaurants pay for that? No. Restaurants can barely afford the delivery fees of people with cars. Maybe Michelin Star restaurants can afford a bit more, but not that much. > Allows medical supplies to get there in a surge. Much easier to just shove it in the back of a C130J. And it requires us to invest a grand total of $0 in new infrastructure or hardware. > I’m sure you can think of something. Again, if you’re going to advocate that it will work, that’s your job.

richardw 6y ago

Let’s think it out (as in: add your thoughts). If it’s for goods, you don’t need a human size pod. You can scale pods to a size that represents goods that make sense to transport quickly. Obviously you want benefits that come from standardisation, like containers give you. Fully automated transport between all major cities for anything less than a metre diameter? I think that looks like the future. Mail, Amazon, fresh produce, run as a stream rather than in large batches, automatically routed, staying off highways. More efficient stock management that can react to surges in demand. Charge for it with spot pricing so high value items get priority. Edit: produce sounds a bit silly but taken to extremes might be interesting. When you deliver a ton of lettuce, it’s sold as a stream. Someone gets the fresh stuff and someone gets the three-day old stuff. If you can reorder every 30m, everyone gets fresh. Smaller shelves and stores, maybe no store except online.

ashtonkem 6y ago
There are two huge problems here: First, very little freight is latency sensitive. The vast majority of cargo is cost sensitive, which is why most bulk goods coming from abroad arrive via cargo ship rather than airplane. There are a ton of companies that’ll accept days or weeks of delay to reduce overall cost; so a loop must compete with rail freight and semi trucks on cost. Theoretically a cargo hyper loop network could compete with domestic air freight, but I find the idea that air freight alone could cover the cost of a hyper loop network hard to buy. This is doubly true when you consider the fact that air freight shares costs with commercial passenger travel, both in terms of infrastructure and vehicle development. Second, they’re already working on automating existing railroads. There’s no particular reason why a hyper loop is necessary to automate a train, and we’re already seeing some of the very long haul train routes in Australia being automated. This comes with the benefit of much lower development time and cost, since it allows for the reuse of the most expensive bit of infrastructure: the rails themselves.
- richardw 6y ago
  I never mentioned bulk goods. I mentioned items that are time sensitive. Most e-commerce is latency sensitive. I recently started getting groceries on an hour delivery. It’s a game changer. Amazon could probably power a network like that by itself. What would a 2 hour Amazon delivery to most cities do? What markets does it enable access to? How do you compete with that? It’s too easy to think of why something won’t work. Rather think “is there a customer and a product for which latency is important but is unfulfilled by current infrastructure”. Don’t use my examples, make your own. Saying “no” to that question is to bet against human desires and ingenuity. Freshest (anything) from 200km away within 2 hours. Will top restaurants pay for that? Latest widget from a warehouse 500km away in 2 hours. Allows Amazon to compete with local stores. Allows medical supplies to get there in a surge. I’m sure you can think of something.
  - ashtonkem 6y ago
    > Don’t use my examples, make your own. That’s ... not how arguments work. You’ve got to provide your own examples. I’m not going to do your work for you. > Saying “no” to that question is to bet against human desires and ingenuity This is genuinely a bad way to approach these things, as you’re presuming that it’ll work based on enthusiasm and not evidence. This exact same logic would’ve led you to advocate for flying cars, as you would’ve assumed that ingenuity would triumph over the practical issues that have so far prevented them from working. Ingenuity is great, but market demand actually pays for the bills. And predicting exactly what the market will do is impossible, but given the fact that this will cost trillions of dollars to actually build, it should be possible to predict use cases that’ll bring in billions of revenue. > Allow Amazon to compete with local stores. Amazon is already crushing local stores on price. It turns out that consumers value delivery speed but not as much as they value low cost. A hyper loop might improve the speed of goods between Amazon warehouses (or run fewer), but it would do nothing to improve the “last mile” problem, which is where all the cost is. Drone delivery + a hyper loop would actually be faster; but I’m dubious that this use case alone covers said trillions of dollars of cost; especially when drone delivery from local warehouses would work too. > Will top restaurants pay for that? No. Restaurants can barely afford the delivery fees of people with cars. Maybe Michelin Star restaurants can afford a bit more, but not that much. > Allows medical supplies to get there in a surge. Much easier to just shove it in the back of a C130J. And it requires us to invest a grand total of $0 in new infrastructure or hardware. > I’m sure you can think of something. Again, if you’re going to advocate that it will work, that’s your job.