HN Explorer

Ask HN: How do self-driving cars help Uber?

2 points by bydenville ↗ HN
Over and over I hear that autonomous cars are the key to Uber's successful future. However I fail to understand how that will help them achieve profitability? how is a self-driving car a cheaper alternative to the status quo? As I see it, the current system for Uber pays drivers (per hour or per ride) only when they are driving. Parking coss, repairs, gas, and the time between rides is all unpaid. As soon as Uber owns their own fleet of taxis, won't these other factors quickly eat away at any potential savings? Although I'm sure there is variability in the costs depending on the geography, how does Uber benefit from needing to continually pay for a 24/7 fleet when they have what seems like the ideal system already: drivers on demand. What am I missing?

3 comments

[ 3.2 ms ] story [ 18.8 ms ] thread
mdorazio
I think you're missing that we already have a profitable business model as an example: rental cars. If it were true that the costs you mention outweigh the potential per-vehicle revenue that can be obtained from renting out cars, Hertz, Enterprise, etc. would all be bankrupt. Obviously they are not, so that business model works.

You can think of Uber + autonomous as a rental car fleet with higher average daily revenue per vehicle (and obviously far more customers per vehicle per day). Admittedly, the costs would likely be higher as well, but the bet is that the net result will be higher profitability overall. By removing the driver cost and largely removing the insurance cost, you end up with highly predictable vehicle costs and can manage your business model around that to remain profitable and hopefully get to a near-monopoly position in major metro areas, supplanting traditional taxis, rental cars, and even public transit along the way.

newyorker10011
For a rental agency the fleet is kept as a discrete number of locations. For a taxi service they need to be kept around the city. In addition, how is the insurance cost removed? I see your point about rental agencies proving that maintaining and housing a fleet is possible, but a mobile fleet in a busy metro area seems orders of magnitude more complicated so at least somewhat more expensive...
mdorazio
I'm not following your first point. Autonomous Ubers would be housed in discrete locations near metro areas in off-hours and then roam around during the day or use negotiated low-rate parking spaces, so the location cost would not be much higher.

Insurance rates are based on actuarial estimates of likelihood to cause accidents. Autonomous vehicles will not be on the roads in a meaningful way until they are significantly better drivers than humans, thus insurance rates will drop low enough that fleet owners just self-insure to cover incidental damage from no-fault causes.

I'm also not sure what you're getting at in your last point. Uber has already built out a highly capable dispatching service to manage large fleets roaming around urban areas and they have a massive dataset to be able to predict ideal fleet allocations. The only real added cost will be the fuel overhead to support cars driving around looking for fares (currently that cost is eaten by human Uber drivers).