[HN Gopher] Demand and ETR Forecasting at Airports
___________________________________________________________________
Demand and ETR Forecasting at Airports
Author : riadsila
Score : 19 points
Date : 2023-05-17 11:11 UTC (3 days ago)
(HTM) web link (www.uber.com)
(TXT) w3m dump (www.uber.com)
| jethkl wrote:
| The post lacks performance metrics, so it is impossible to know
| if their effort was worth it. It would be improved by explaining
| how performance is assessed and why their method is better than
| an existing or much simpler (e.g., rules-based) baseline.
| zamnos wrote:
| > One way to tackle this undersupply/oversupply issue
|
| ...is to pay the drivers while they are working for you. While
| waiting for fares, they're not free to do whatever, and should
| thus be compensated. That Uber won't pay their drivers while
| they're working seems criminal.
| rippercushions wrote:
| Are there _any_ taxi-like systems anywhere in the world that
| pay drivers for idle waiting time?
| zokier wrote:
| In finland taxi drivers are regular salaried workers earning
| hourly wage during working hours regardless if they have
| customers or not and they may earn additionally provision of
| sales. Caveat here is that some taxies are self-employed
| which makes the situation bit more complicated
| 2Gkashmiri wrote:
| uh, pay them a monthly wage for doing 9-5 and the drivers
| would drive as much or as less, what the app makes them work
| tsimionescu wrote:
| The question stands: are there any taxi-like systems that
| work like that? Traditional taxies certainly did not employ
| their drivers and pay them a wage, they were still very
| much working for their fares.
| bombcar wrote:
| Hired cars sometimes pay their employees that way (the
| driver gets $/hr no matter what), hence you have to
| schedule well in advance.
| zokier wrote:
| Hypothetically if drivers were salaried, wouldn't uber have
| even greater motivation for implementing this sort of system?
| oh_sigh wrote:
| If that were the case, droves of drivers would just head to the
| airport to relax in their car and watch youtube videos.
|
| It would require uber to actually tell drivers where to go,
| which it doesn't do right now. Basically, completely change the
| model currently used.
| bombcar wrote:
| How useful would it be to Uber for me to tell them I'm arriving
| at X and will use Uber? Enough for a small discount?
| drewg123 wrote:
| How about getting estimated wait times for pickup for _riders_
| even remotely correct? That helps drivers too:
|
| Recently I had a severely delayed flight that didn't land until
| after 1am, and was by far the last flight into my podunk airport.
| My apps showed a 3 minute pickup time from Uber and a 20 minute
| time from Lyft, so I booked a ride with Uber. After 3-5 minutes
| of "this is taking longer than expected", it finally found a
| driver, and the pickup time wound up being 20 minutes (just like
| Lyft). So I cancelled and got into a waiting cab and was home by
| the time the Uber would have picked me up.
|
| If Uber would have just been honest about the pickup time, I
| would have never booked a ride and would not have wasted some
| poor driver's time when he got the ride request. This also
| delayed me by 3-5 minutes (the matching time on the app), which
| irritated me. I'm gradually being conditioned to not trust Uber,
| and to just use Lyft and Taxis..
| totoglazer wrote:
| This is such bullshit. I'm sure they A/B test and find that
| outright lies helped with whatever metrics. However I too don't
| trust the apps, including Lyft, to represent pickup times at
| all accurately. That's assuming good faith on the drivers,
| which I've found to be very dubious.
___________________________________________________________________
(page generated 2023-05-20 23:02 UTC)