The evolution of transportation-as-a-service
stratechery.comOne guess: the routing problem(the vehicle routing problem) is a core computer science problem, for decades. Tons of research have been done about it. I don't expect Google or Uber to have any breakthroughs there, because it's pure luck. But maybe they will, or maybe they'll will ackuire someone. Let's put it down to pure luck.
But assuming no brekathrough happens: one company that has tons of experience in that field is UPS, they are just finishing building a new system for running their vehicle-routing for 50K routes in the US daily, ~120 points per route, i.e. 600K customers daily. They invested ~$300 million in that system(and they employe hundreds of phd's) and they say it will give them a savings of less than 1% of revenue - just to get a sense how mature those systems are.
On the other hand, it seems that have larger access to customers could enable much more ideal routes, both in cost, customer time and maybe social factor(if social matches will be part of ride sharing). So i believe that to be key to winning.
And i don't think anybody can beat Google in marketing to android users, and even in the US, with iPhone's 40% share(but more wealthy users) - that's avery big advantage for Google.
On the other hand, i think Google prefer not to start a huge service that employs many people, and than have self-driving cars make them unemployed and suffer the huge reputational damage. So they won't go into true shared taxis like ridewithvia.com seems to be doing very sucsessfuly, and instead stick with wazer-rider , enabling drivers to give a lift to poeple for some very modest fee.
So to a certain extent, the winner in that battle would be determined by which approach of those two will win.
For this problem, they don't need to find the optimal routes, just reasonable routes. Individual routes have only four stops at a time. No breakthroughs required.
Not choosing optimal routes would lead to a more expensive service, which is critical for a commodity.
As for routes only having 4 stops - sure, but that's after you chosen which users will drive in the same car/trip, which is hard in itself.
BTW the dial-a-ride problem is quite similar to the ridesharing problem, and the complexity there is O(number_of_pickup_and_drop_points^2 ) [1]
[1]https://www.itu.dk/people/pagh/CAOS/DARP.pdf - altough it's a bit old, so maybe results have improved since.
If you compute routes that are close enough to optimal, it doesn't matter. If you compute good enough routes quickly, that will likely reduce costs more than computing the best routes slowly. There are enough other areas of optimization that will give a better return on investment that it doesn't necessarily make sense to invest in finding the actual best routes.
Also, this particular problem is much easier than you surmise. The problem statement isn't a set of millions of cars sitting in particular locations with a set of millions of riders with particular sources and destinations and matching them and ordering routes to minimize fuel usage or median rider travel time. The problem is of hundreds of nearby cars traveling existing routes and matching a single new rider with a particular source and destination to one of those cars. There is a secondary problem of where to send empty cars that is more interesting.
I agree. But that's true for current Uber .
But once we start talking about UberPool and via its pretty close for the problem I describe, especially if we're talking about the commute , with a possibility of people preordering rides sometime in advance(or even some prediction abilities about that).
As for the other option, carpooling ,it depends how far will people be willing to drive out of their way for that extra income. But being conservative and I they won't go out of their way , the problem becomes assigning riders to a million bus routes with small capacity, which seems to naturally break this problem into many ,largely independent problems , and may prevent any big exponential complexity.
Don't you agree ?
Uber actually transports people.
Ford actually builds cars.
Google... likes letting smart people do smart things.
While Ford may be the underdog in the race, I like its chances. It has the operational experience to take a "works-in-concept" to "works-in-reality". The software part, while difficult, is not the hardest part of "TaaS". The systems part is.
I agree the systems part is the hard part. The software to monitor and control thousands to millions of automated cars.
Ford has a large learning curve in order to figure out how to do that. It will take a very different workforce to build and manage a software system of that scale.
Google and Uber have been managing a system like that from the beginning, it is in their DNA.
I could see Ford providing the hardware, while a different company deploys and manages the customer facing operation. Demoting Ford to a vendor doesn't sound like such a good thing for Ford though.
The race also includes GM, Tesla, Faraday Future (underestimated, but extremely well financed), Zoox- a stealth unicorn and Apple, among others.
While we'll see autonomous taxis operating in some capacity soon, they have to work extremely well and be profitable to justify an aggressive rollout, and that's likely more than a decade out. It's anybody's guess as to what the playing field will look like then.
Many analysts and speculators are getting ahead of themselves by conveniently ignoring just how much of the technical problems of autonomous vehicles need to be solved before these vehicles are ready to venture beyond very carefully managed services operating only in optimal conditions.
The takeaway I get is that Uber has an edge on Google in the race for a fully automated taxi service, because the have had more time to work on a more sophisticated routing algorithm.
I am skeptical. As the article points out, it will take a considerable amount of time to deploy any service, once the automated driving is good enough. If Google gets to commercially viable automated driving sooner, it will have plenty of time to refine their own routing algorithms during the rollout.
An interesting HN comment on vehicle routing from yesterday:
https://news.ycombinator.com/item?id=12400611
My sense is that this problem is solved to a much greater extent than automated driving is.
Uber has the advantage because they have been "Doing things that don't scale" and have a large knowledgebase of how to effectively operate a taxi service, with customer info/ratings as well. User experience trumps routing algorithms.
Hasn't transportation been a service you can buy for ages?
Yes— but it hasn't been cost-competitive with a driver-owned car for daily drivers (given free/subsidized parking and roadway space, as we have in the US).
Yes, but it's changing, which is why the article is about the evolution of that service (and not, say, the creation of that service).
I would be surprised if the planning part is the most difficult part of autonomous cars. Embodied intelligence is really nowhere yet. I've seen Google using particle filters to represent other cars.
AI is just not advanced enough to cope with leaves on the road, a pedestrian who wants to cross or not, a broken traffic light, a criminal who want to steal your car, etc.
And everything as a service... Really, if we would have something that awesome as an autonomous car, wouldn't we want to own it!?
I would! I want to talk to it. And I would like to have a bed and a bath in it and have it go on a road trip with awesome beaches, castles, and sunsets.
> Really, if we would have something that awesome as an autonomous car, wouldn't we want to own it!?
Because money. Let's say the first autonomous car is a Ford Fusion or similar, normal cost $30k, autonomous cost $130k (for the sake of argument; we know autonomous tech will be very expensive at least at first.)
Would a normal person buy the autonomous one? Not by a long shot.
But for a taxi company, saving as much as $100k/year on payroll expenses per car (assuming shared between drivers), the $130k car is very attractive.
If it will be successful, many will be sold and the price will come down quickly.
I don't know why AI should stay expensive.
iPhones are successful, many have been sold, yet the price has not gone down quickly.
For autonomous cars, the best-case component price estimates I've seen for solid-state LIDARs, assuming mass production, is $1000. I believe you need four of those, and then you need ultrasound, radar and super-hi-res cameras. And a big honking computer to process all that data. It's probably at least a $10k premium over the normal car just in additional hardware costs, and that's assuming mass production.
Then there's the cost of all that software development that companies will want to recoup, plus the cost of collecting and live-updating a vast data set of roads, pedestrian paths, local wildlife/other hazards, roadworks, local laws and regulations etc. in every big city and small town in all the world, to assist and confirm the information gathered by sensors in the car.
Solid state should be cheap.
The real problem is that LIDAR instead of better "stereo camera data processing" is development in hardware rather than brainware (AI).
There is indeed a lot that needs to be developed w.r.t. brainware. However, I'm pretty sure it doesn't need to be recouped through selling cars. Brainware is valuable in many more applications.
1. Accessing the internet through Google vs Viv. 2. Household robots vs cars. 3. Agriculture.
iPhone prices don't go down because it's a brand and hardware controlled by Apple. AI is software, with zero marginal cost to copy.
Owning costs a lot. You need space to store it (in your house, near your job, near stores). You need to maintain it. It sits idle most of the time. Some people think it's worth it to try a different model.
True. It's cool.
You what else is cool ? an RV. everybody dreams of owning an RV. almost nobody does.
An autonomous RV, now we are talking. :-)
So it will vacation for me, while I'm working? Or just drag me on a best-fit itinerary when I'm aboard? Either way I don't envy other road users, I already feel like RVs are hulking, slightly unpredictable monsters when I have to drive or cycle around them.
About the only modification I would add to TaaS is making ride-sharing optional -- a personal preference of mine, that might prove worthy of a premium. If I liked sharing space with others, I would take a bus to work.