Does anyone know if Lyft plan to contribute their map corrections back to OSM? They mention other companies doing so in this blog post, but don't make it clear whether or not they, too, intend to.
I'm curious to see how good Lyft Maps is, too. Uber Maps is, frankly, a disaster in my area, frequently suggesting illegal and/or impossible turns. I try to submit map feedback every trip (very hard as a rider, as most forms of available feedback seem to be to punish the driver rather than the platform), and it's still never been fixed. Most drivers seem to use Google Maps or Waze instead and if I see a driver using Uber Maps, I need to watch the trip and suggest several corrections. Ironically, I'm only a few miles from the office where most Uber Maps staff used to work.
I also find it consistently obnoxious that there is essentially no way to say "My driver was fine, but your platform did something stupid" when leaving feedback. At least, not without jumping through a million hoops. The arrogance of our industry is tiresome.
I took a ride in SF a while ago and it predicted a 3 minute pick up time which turned out to be 25 minutes - no fault of the driver, and the only way I could complain about this was to click through like 4 levels of "other issue" and then jump through hoops sending emails back and forth with support.
Pick up time predictions are garbage, because ride shares allow cancelations. Last time I used a ride share, I had eight drivers accept and then cancel on me. The predicted departure time was always "8 minutes away" and then "looking for another driver" over and over. I probably waited an hour.
Talking to a driver about this I was told that the platform is always re-routing drivers to maximize revenue for lyft, so when the app says "finding you another driver", it's not the driver that canceled, it's the platform decided that driver was better off, revenue wise, to complete a different fare.
The fact that this is opaque to the end user just means we get mad at the workers and not the corporation, classic.
The fact that OP can see "finding another driver" 8 times suggests that perhaps there is some prioritization system, and OP was low priority - such that if any other request came in before him, then the driver would be re-routed.
It's much simpler than that. Drivers accept and then they cancel the ride themselves if they don't like the destination or the earnings. Normally you would penalize those drivers but I imagine with Lyft losing market share they likely can't afford to penalize drivers all that much.
No, drivers don't see the destination until they arrive and the trip begins.
I'm often on one side of the Bay Area trying to get to the other and had many drivers pull up, then cancel or try to get me to cancel when they see I'm going across the Bay.
That's at least not true for Lyft as of October last year. But there are many factors, like other apps, traffic, getting offered something better from another app, etc.
> With upfront pay, drivers can now see ride information and what they’ll earn before accepting a ride. Drivers will have access to trip details including the pickup and drop-off locations, estimated time and distance to complete the ride, and a map view of the full ride in addition to the fare.
My assumption was that most drivers are using 3+ different apps, so if they get an Uber request while they're driving a Lyft route they just don't accept it, hence the bouncing around "finding another driver" because all the drivers are already busy
Also, riders are using two or three different apps, so if Uber promises a 2-minute wait and then confirms with a 6-minute wait and Lyft is promising a 3-minute wait, cancel the Uber ride and now that driver is available again.
Possible that those shitty tips you've been leaving are adding up in other ways? I live in a neighborhood that borders a part of town that is accepted as a "bad" part of town and is often confused as being part of that part of town. I have only ever not had the "finding another driver" twice from my house, and originally assumed it was location location location. I had a guest stay with me last summer, and they never received that treatment which led me to believe it was an account thing. We'd test by requesting a car at the same time. Their rates were cheaper, and they never had a new driver issue.
It would make perfect sense to have some ML model which predicts which customers will be most profitable, and treat those guys well to improve retention amongst that client set.
(Regarding Uber, not Lyft, but...) It would also make perfect sense to predict which customers are investigators in areas Uber is illegal, and avoid them. Which they did [0]. I would be shocked if they weren't giving more profitable customers higher priority and better treatment.
Forget ML, Lyft can tell if you are a more profitable / more engaged customer through your usage of Lyft Pink, which explicitly offers Priority Pickup. They are transparent about getting you a driver sooner, for cheaper, and I guess the subscription you pay for drives the value.
It’s especially rich when Lyft gives you this treatment after you’ve selected the more expensive “rapid pickup” option. Nothing hits quite like the tech company gaslighting experience of being told “Just a moment…Just a few more seconds…” for upwards of five minutes by an app eager to discourage you from canceling your $5 extra “three minute pickup.”
This boils me to no end. Google Maps estimates often make rideshare look like a timesaver compared to transit, but when you add five, ten, even fifteen minutes to the pickup estimate, it really changes things.
I had a ride from YVR airport where no less than 30 drivers[0] cancelled on me. I would get the notification that the driver was on their way, tap it, and then catch a brief one second flash of the driver before they disappeared and I was back in the queue.
[0]: I’m sure it would be much more, but I gave up and took the train instead
i recently flew to Austin a few times and the turnaround time from airport to destination using waiting taxis was far, far faster. not cheaper though! i have found these Uber/Lyft lots that airports are using now to be a bit of disaster; they have markedly reduced the convenience factor of uber/lyft from airport to destination.
Do they actually take credit cards nowadays? And do they expect gigantic tips?
I really hate shelling out $60 for a 20 minute taxi ride, then being expected to shell out an extra $20 on top of that and getting endless grief over wanting to use a credit card to top it all off
Every cab I've been in in the last decade has expected that you're paying by card. Many prefer it, because the drivers don't want to carry a bunch of cash. Many cab companies even have apps where you follow the progress of the cab you called, and pay for the ride.
I have no idea what the tip expectation of cab drivers are. I tip the same rate I tip for all traditionally tipped services. I don't actually care whether or not that meets their expectations.
I've known a couple of people who never tip cabbies. Nothing bad happens to them.
> I really hate shelling out $60 for a 20 minute taxi ride, then being expected to shell out an extra $20 on top of that and getting endless grief over wanting to use a credit card to top it all off
Drivers actually comment on your tip amount? And they complain about an extremely generous 33% tip?? I wouldn't use that cab company ever again.
When I was a teen I would tip the driver whatever the change would be to the cash I had on hand. These would be trips between LAX and West Hollywood. Cost like 25-30 bucks? I don't remember.
The only cab I saw that took cards had a thing that would slide across the front and get an impression of the text embossed on the card. I guess that wouldn't work today since cards aren't embossed anymore
At airports, its often much better to walk just outside the boundary of the airport.
Airports often have special rules, pickup fees, etc. Drivers don't want to pay an $8 fee just to make a loss on your journey. As soon as you're a few yards out of the airport, none of that applies and a driver will be happy to take you.
>its often much better to walk just outside the boundary of the airport
are you serious? what airports do you visit where you can do this as a suggestion without laughing. It's like telling the freshmen that there's a pool on the school roof.
even something as small like a Bob Hope in Burbank would be difficult. Most airports do not consider pedestrian traffic to the edge of their properties. I've done it once at that airport because my office was pretty much next door to the Fry's across the street. still made me regret that decision. I'm now in Dallas and even Lovefield pedestrian traffic past getting to the parking area is non-existent. From DFW? Forget it. Much like your suggestion
Since the thread is about YVR, it's actually pretty easy and safe to walk all the way to Richmond (or a shorter distance to the east side of Sea Island which has normal businesses / residences). Sidewalks all the way from the terminal, and no crazy crossings are required. I'm sure this isn't typical, but if you don't mind the walk it's perfectly viable in this case.
You can also take the SkyTrain for free within Sea Island (which is not well known, I feel) - 2 stops to Templeton, and there are probably plenty of Ubers around there, since there is an outlet mall near that station, or at least a much shorter walk into Richmond.
LAX: walking paths and signalized crosswalks get you all the way to the hotels east of Sepulveda, or take the free shuttle to the cell phone lot and order from there.
SJC: Cross the parking garage to the bike path and from there cross the street. Or just walk down the bike path
ORD, MDW: Free shuttle to the rental car facility, or take the L one stop.
MSN: kind of a long walk, but there are sidewalks the whole way
SLC: take the light rail one stop
PSP: dunno why you’d want to walk in the heat, but it’s a short walk on sidewalks to El Cielo Road
Bear in mind: The minimum wage retail staff that run all the stuff in the airport can’t really afford to pay even economy lot parking rates, so basically every large airport has some kind of public transportation if you know how to look for it.
When I fly out of Seatac in the summer I park my motorcycle in the employee parking, which is free. Do other airports not offer cheap/free employee parking?
Walking out of an airport of any size (and even many/most smaller commercial airports) is between impossible and taking your life in your hands given that sidewalks/crosswalks/etc. are largely lacking. Most commercial airports just don't contemplate pedestrians walking in and out. It might be possible to walk to East Boston from Logan airport though I'm not sure how--and that's an airport that's basically right in the city.
There's a sidewalk from Terminal A arrivals that you can take to the Maverick neighborhood and the East Boston Greenway. There was one from Terminal E as well to Wood Island and beyond, but part of it's been closed for a while due to the Terminal E expansion. There's a pedestrian detour through the cell phone lot that'll get you there.
You won't be able to walk to Downtown Boston, obviously, but Logan Airport is theoretically walkable if you're in East Boston. I'd still take the shuttle bus to the Airport T station where you can do the same walk to the greenway.
Most airports have moved to a centralized rental car location that is still inside the airport perimeter, so not sure how this is helping? The ride share algos are still going to recognize you are at the airport and charge you accordingly
Logan is one of the easiest. You just take the bus to the MBTA stop or the free silver line into Seaport. The convenient public transportation is the reason I rarely take cabs or rideshare to or from Logan.
Sure. It's convenient to public transit. Really no reason to catch a cab to downtown unless you have a lot of luggage, you're going in or out at a weird time, or home is somewhere inconvenient to public transit. I was responding to literal "walking." (I live outside the city personally and transit doesn't really work.)
It is very easy at LAX. You walk about 100 yards across one (admittedly very busy) intersection and get picked up for 1/3 the price. You already have to take a shuttle to even get picked up unless you have a black car, so this adds 5 minutes. It is still annoying if you have luggage.
You could also take a shuttle to a local hotel or car rental, but that is breaking some rules.
What 100 yards are you talking about? Stepping out the doors at any terminal at LAX takes you to the parking lot if you walk 100 yards. To get off of airport grounds would take a much significantly longer walk without any pedestrian walkway
I am talking about the walk from LAXit. This is the parking lot where Uber/Lyft/Taxi pickups are, and you generally have to take a shuttle to get there to begin with.
LAXit is on Sepulveda, but it is safer to cross the street to a parking lot. I'll admit I underestimated the distance. It is closer to 300 yards:
I think for users who comparison shop between Uber and Lyft, the pickup ETA is part of each app’s bid. Putting a lower price costs them money, while putting a lower ETA is free.
Which is why Lyft does that annoying 59-second countdown until you can see your driver's details and actual wait time, so you have a sunk cost to discourage you from switching to Uber when the actual wait time is longer than promised.
This is a case where a positive AB test result could push you in the wrong direction. It might take a long-term holdout group to find the negative effect, and those can be hard to manage.
Sometimes, you'll report a minor issue and get an over-the-top "OMG - WE'RE SO SORRY! HERE'S FREE MONEY AND YOU'LL NEVER BE PAIRED WITH THIS DRIVER AGAIN!"
Other times, you'll report an actual problem, like being charged more than advertised, and they'll do nothing to remediate.
After being charged more than expected on a couple recent rides, I basically stopped using Lyft.
If it's anything like AirBnB, it probably depends on whether you're getting the chipper day-shift call center or the cheap outsourced offshore call center as far as what kind of support you get. My wife and I landed in a cockroach-infested but very beautiful condo through AirBnB once and the night shift guy basically accused me of being used to living with cockroaches as a business traveler and that one or two should be expected. I had to offer to send him video of me spraying the walls of the kitchen with roach spray with tens of wriggling bodies of various instar stage falling onto the counter before he would even commit to looking in to it.
I called the same support number the morning after and the support people were all very accommodating. Called them one more time that afternoon to confirm that our payments were refunded and got the same treatment as the previous night.
This experience is one of the reasons I never use AirBnB for anything anymore. Their support was awful at the exact moment I needed them to offer me any option other than sleeping in our rental car on the beach.
I reported the space being poorly cleaned first thing in the morning (I had arrived at 1 AM). They asked if I preferred a phone call or chat. I didn't want to spend a morning on the phone with them, so I chose chat.
Every 20 minutes for the next few days, I got "Hi, I'm NAME_HERE, and I'm taking over your case!" The names changed, but they never actually _did_ anything beyond "taking over" the case. I felt like I was in a Seinfeld episode.
One of them then decided the best resolution was to tell the host I'd complained about the cleanliness (without including me at all). I started getting nasty messages from the host, who thought I'd happily stayed there for 4 days and then spontaneously opened a complaint. So now on top of sleeping in a gross Airbnb, I was getting constant notifications on my phone of the host calling me all sorts of nasty names.
I opened a _new_ ticket and chose phone support to finally get a hold of somebody. Her accent told me she was in an offshore call center, but she was helpful in getting it resolved. I think they had me book a hotel for the remainder of the days and reimbursed me, or something to that effect.
After I checked out, the host filed a frivolous damages claim against me. I clicked the "not my problem" button and thankfully never heard about it again.
I will never, ever use Airbnb chat support again. Was perhaps the worst customer service experience I've ever had.
After ten years of using Lyft/Uber for airport trips instead of driving I’ve decided to go back to park/fly companies. This past week I had a scheduled and confirmed pickup with Lyft, driver cancelled 2 min before pickup and it took 5 more “driver found” messages and 50 minutes before I switched to Uber and got picked up in 10 minutes. Uber has done the same in the past, so they aren’t reliable either. It was a $110 fare trip.
I’m beginning to despise Lyft/Uber as much as the cabs they replaced.
This - I find 2/3rds of my issues are platform vs driver. And there is no way to penalize the platform vs the driver + the driver then deals with the aftermath of a pissed off customer even if they were great. It would be as easy as having two star ratings, “how was your driver?”, and “how was your whole experience?”.
Totally agree, I just had this happen to me two days ago. The Uber app bugged out and left me with a fare that was double what I was quoted, and there was basically no way to get in touch with Uber about it. You know how often support chatbots offer a "talk to a human" option? Not Uber. Just links to articles about how I'm holding it wrong.
It’s worth taking to Twitter to complain in those situations. I’d love if we didn’t half to, but I’ve found most companies help when you make your issue public.
Maybe Uber isn’t a company who will care, but most are.
Lyft is one of the biggest contributors to OSM. Here is a list of OSM accounts related to Lyft [0]. In many places where I've edited OSM, Lyft employees have contributed quality data where there wasn't any before.
How (and, how respectfully) companies should contribute to OSM is an ongoing debate. But corporate influences are often a good thing, and I think nearly anyone would agree that Lyft's contributions to OSM have been resoundingly helpful.
Isn't this the driverless car team, vs. the ride-mapping team?
Regardless, it's good to see that they are contributing back to OSM. I agree that overall the use of OSM by anyone "lifts all ships" as long as users continue to contribute back. OSM is one of the most positive projects around lately in terms of taking something that used to be extremely expensive and highly restricted and democratizing it.
That may be true. But as a whole, it is worth noting that Lyft hires a significant number of people to edit OSM. Here is an interesting 2021 post by Jennings Anderson discussing paid edits on OSM, and how companies compare in their contributions:
I agree about how successful OSM has been recently. In an age where Wikipedia for example seems relatively mature, it's fun to see OSM grow at such a pace.
I used to work for Lyft and I can confirm that both teams (driverless and ride-share) contribute. Ride-share scaled later which may give you that false impression.
And Yes, it's a big investment for Lyft for the open source community.
Yes - Lyft contributed the overwhelming majority of the new data back into OSM. The only exceptions were layers that didn't meet OSM community standards for various technical reasons.
They're saying "we contributed the overwhelming majority of the data we collected and corrected, except for that which wasn't eligible for contribution." It really seems like Lyft operated in good faith here, they just didn't mention it in the blog post.
I've added sidewalk vendors to the map with names like "coconut cakes lady," and had them approved instantly.
I've also tried to rename streets, including the news article about the name change, and had them unceremoniously ignored. Same for adding a new bar to the map with a photo of the menu, even though other places I've added have had hundreds of thousands of views.
There doesn't seem to be an underlying system. There's a famous traffic stop in Muine, Vietnam listed as something like "social security office." The photos are all of the traffic stop. The reviews are about the traffic stop. If you run the name on Maps through translate, it comes back "traffic stop." I've tried repeatedly changing the category to "traffic police" and been silently denied.
There's a bar in downtown SF at the wrong address. Comments are about the address being wrong. There's even a photo that says "our address on Maps is wrong." Try to fix it in Maps? Denied.
Providing feedback to large corporate machines is akin to screaming into the void at this point. Sometimes the void hears you and responds with a generous canned message. Other times you unsurprisingly get silence. That's why I became completely divested and disinterested in providing general feedback for maps, locations etc.
Map creators include intentionally incorrect information (Map Traps) so people copying them are obviously copying them. I sometimes wonder if that’s the root cause of companies failing to correct some of their mistakes.
Using anything that could cause incorrect directions as a trap location / trap street would be a ridiculously bad idea. I strongly suspect the reason is a more standard corporate one; the volume of corrections and signal to noise ratio is probably quite poor.
In the case of my repeated Uber map correction attempt, I suspect that my correction isn't prioritized because of two factors: first, the turn which Uber Maps suggests, while both illegal and very dangerous, is not impossible (it's an illegal left turn accomplished by making a U-Turn around an island designed to prevent left turns). So, they probably see enough volume of "driver completed route as suggested" data signals to ignore any "but it's illegal and dangerous" complaints, a classic problem with "data driven" systems. And, Uber don't seem to have a real process for riders to submit mapping corrections, only drivers, so my request is unlikely to ever be routed to the correct queue to begin with.
Map Traps (as these are called) are normally pretty harmless, e.g. a non-existent, small cul-de-sac that doesn't exist. A turn restriction (which is quite invisible!) seldom is.
Meanwhile, in Toronto, I see some bigger driveways listed as roads on Google maps, and I always wonder if it’s a legally correct city survey artefact and I could park a car there etc. if I wanted, or just a trap/misclassification.
The way to get that corrected is to ask the local police to ticket illegal turns at that intersection for a day, indicating that Google refuses to correct the issue and is directing hundreds of violations a day at the intersection. Google ignores unenforced road signs that are frequently ignored by their data providers (drivers), so enforcement complaints from drivers are necessary to force the correction.
Here is an overview of which corporation contribute how many edits based on the osm account names affiliated with the corporations: https://piebro.github.io/openstreetmap-statistics/#b34d Lyft is the eleventh most active company by edit count with a total of 4M edits and 1.5M edits in the last year.
> By late 2019, Lorimer and Justin Moore, then the head of engineering, were ready to form a team to “do the impossible” — create a home-grown mapping system.
THE IMPOSSIBLE: A "home-grown" mapping system built on OpenStreetMaps?
Lyft seems to really lack any imagination if _that_ was considered impossible.
If the managers and ICs agree that something is impossible but they will do or have done it anyways, then that makes a pretty good case for them getting promotions.
I have thought that some major service migrated to OSM by 2019. I can’t remember which, or even if it was a success. But I didn’t realize the data was that bad just three years ago.
Then again, Apple Maps started getting sorted out around then too iirc.
> Then again, Apple Maps started getting sorted out around then too iirc.
And we assume completely coincidentally? Not that Apple Maps, like Google Maps, pulls in data volunteered by the OSM community (both corporate and personal members), packages and improves (sometimes "improves") it, then locks it away behind paid API calls?
In this case and many others, the free and community built project enabled, assisted, or powered the multi-billion dollar tech conglomorate.
I recognize I might sound bitter or dismissive of companies mapping the world and giving it away to consumers for free, but so much is built on openness that so many companies just eat without giving back.
> Not that Apple Maps, like Google Maps, pulls in data volunteered by the OSM... then locks it away
You have that backwards, at least for some companies.
> How Facebook, Apple and Microsoft are contributing to OSM
At the State of the Map conference in Milan, the teams from Microsoft, Apple and Facebook presented their projects, describing how they are working with communities.
> You have that backwards, at least for some companies
Absolutely I do - my overall (and apparently poorly communicated) point was more about Apple pulling in data from OSM to improve their product. They (and others) do contribute back, I only meant it's hardly coincidental that OSM and Apple Maps improved around the same time.
> I've never seen any cite for Google contributing their data to OSM.
This is what I'm so up-in-arms about: people spend so much time adding to Google Maps for free (not even necessarily via OSM), to be locked away behind paid (past a limit) API calls. Perhaps I should have been more explicit in that many companies contribute back, and Google being an outlier in that.
> it's hardly coincidental that OSM and Apple Maps improved around the same time
True, but in Apple's case, the massive improvements stemmed from no longer relying solely on data licensed from others, and starting a program back in 2018 to collect their own map data from scratch.
Apple Maps doesn't use OSM for developed countries, but for places like SEA and Africa. It's places like Facebook and Mapbox who use OSM for everything IIRC (don't trust me here as I didn't check this.)
Google dont directly use osm, as far as we know, but they or subcontractors check the changelogs and often add newly built things hours/ days after they get added to osm, presumably from their own sources as the data is not identical.
> Lyft seems to really lack any imagination if _that_ was considered impossible.
Depends on how you are making the maps. If you're paying someone to trudge around/trace out satellite photos, then yes, its not very imaginative.
If you are using SLAM to generate maps automatically, then thats a different beast entirely. Only a handful of companies were able to do that in 2019, two got bought by meta (who then fucked up the tech, well done zuck) two went bankrupt and one was aquired by lyft.
it's not about not paying Google Map licensing. There are, as mentioned, geo features like pickup ETA, trip ETA and cost, drop-off location photos and features that Google doesn't provide at all.
This is yet another "We found a community project providing us with billions of dollars in free work but don't find it worth the effort to contribute anything back"
Big corps that use OSM are actually contributing a lot back into OSM by actually fixing data issues. I have seen that first-hand; at a Corp that was using OSM, they had a big team doing just that the whole working day. Just fixing issues in OSM that the customer have reported.
I think they in the end decided to fork OSM and make it proprietary to keep the data in-house; but I might be wrong. I was not there anymore.
edit: maybe I am wrong. It was Grab, "uber for Asia". They still edit OSM it seems...
Looking at all edits done by osm member affiliated with Grab from that wiki list (https://piebro.github.io/openstreetmap-statistics/#b34d) it seems they contributed a lot of edits (21M in total, sixth largest company contributor). The amount of edits slowed down on the last year though.
Honestly you would think Lyft would be more explicit about contributing back in their article.
This isn't the first time I've seen a company contributing back to open source but completely neglecting to mention it in their post about it. It's weird, usually companies love to virtue signal about any tiny thing they do for society.
> Share-Alike: If you publicly use any adapted version of this database, or works produced from an adapted database, you must also offer that adapted database under the ODbL.
I really wish that companies had a donation program where employees get to vote on their favorite project, and the company cuts a check every time a project hits 50 votes (including things that got 20 votes each for the last 3 election cycles)
We have something like this at our firm (IT consulting shop). Every Christmas, everyone gets to propose & vote on projects (software and otherwise) that we should fund.
Oof, I'm afraid I don't know of any. Our company is generally very, let's say, employee-oriented and has been doing donations this way for a long time (longer than I've been there), so I don't know how our bosses came to like this approach.
One argument would certainly be: Employees / engineers tend to know much better which projects are worth supporting and/or in need of financial help.
I think GP is saying they expect Lyft respond to complaints or things to change by fixing them in their internal OSM database, but maybe not propagate those changes back to canonical OSM for everyone else to enjoy.
Couldn't that become hard to maintain after a while though, once the OSM map changes and they start getting conflicts? I think that's often the biggest reason why corps contribute back to open source projects rather than maintaining internal forks for everything.
I guess it's possible though they could just deal with that if they feel like having better maps is a significant competitive advantage, but if the goal is just to not be dependent on Google upstreaming the edits sounds like a better idea.
I think both sentiments can be valid. If OSM really needs those contributions they need to take responsibility for updating the license. But we as a community should encourage companies to give back to the OSS projects they benefit from.
Well, it is not as if the OSM-community is a monolithic one. Some people are hardline anti-corporate and don't like this; others are part of a big corporation and see this as business as usual.
Even if Lyft would not give back in some form (spoiler: they do, see https://wiki.openstreetmap.org/wiki/Lyft), it spreads awareness and drivers will figure out they can fix issues themselves.
> Ninety-eight percent of drivers who try Lyft’s maps stick with them — rather than switching to Google, Apple, Waze, or any other app.
I find this surprising because my personal experience is the exact opposite. At least in NYC, drivers use the Lyft (or Uber) app for managing riders but use a dedicated mapping application for navigating to the destination once you're picked up.
Agreed. A lot of drivers in NYC (where I'm also based) seem to have multiple devices in the car. Maybe there's some flawed data gathering if a driver still is using the native Lyft app but running navigation on a separate device?
I'm sure they take into account whether the driver is following their app's suggested route. If the driver disregards the suggestions that would suggest the driver isn't using their maps.
Lyft and Uber have been very disappointing at the local airport lately. Lyft pickups from the airport are supposedly cheap but the last time I tried getting one it let me wait for 20 minutes; I eventually gave up and got a taxi.
Uber from the airport regularly surges, usually right after a flight lands, and their surge price puts them well above the price of the taxis that are ready and waiting to pick people up. I’m not even sure why anyone would take an Uber at surge prices - and it’s not like their regular price is that much cheaper than a taxi anyway.
Also, the taxi companies have gradually been getting better at the app game. They’re still a ways off from Uber/Lyft’s usability, but they’re also less likely to be flaky and they never surge.
I compared pricing from hotels next to the airport (that have free shuttle service) and they appear to be $20-30 cheaper on Uber vs the airport pickup point. Planning on trying this on my next trip.
Good for them (and I say this as someone who was actually in Google Maps, both in monetization and in regular development).
Maps has been losing money for literally decades. I don't know if they've finally turned that around or not. In some parts of the world, they were still paying a 3rd party per lookup because they didn't own the data.
I also don't know the numbers for their API charges, but I wouldn't be surprised if the totals are mostly symbolic, for VP's to defend themselves as "Hey, look, we're getting all we can out of it!"
Cool startup! I did some work for a client in the trucking space a while back, and I came away with a new appreciation for what a fascinating vertical it is.
If you don't mind me asking, how'd you end up in the space? It feels like the vertical has so much esoteric/specific knowledge. We ran into a few companies that were founded by truckers (or by people who had truckers in the family), and it seemed like already having all this knowledge gave them a big leg up.
First, the focus is on savings to Lyft, and not how this innovation benefits customers in any way.
Second, it is highly antagonistic to Google - who is probably still a partner to Lyft in many other aspects (advertising etc). Taking shots at your partners is generally considered bad for business.
Savings benefit Lyft customers directly by keeping the company solvent – since the rides aren’t being heavily subsidized, the rates are going to go up and things like this will reduce the sting.
It also helps indirectly by giving them control. Google doesn’t do product development well and Maps has been stagnant or even regressing for years. Lyft can always flub it but now they have the incentives lined up where the only party needed to make what Lyft users want is Lyft.
As for Google, they’re a big company and they can take criticism. It’s certainly not like they’re shy about canceling products or changing the terms any time some MBA thinks it’ll help their bonus, and I’d be quite surprised if they didn’t pass on the opportunity to keep a customer multiple times before now.
Cost savings was definitely a factor, but this is the key point from the article:
"And yet, until very recently, Lyft had limited control over that navigation experience, which was built using Google Maps.
“For our entire existence, we couldn’t fundamentally change that product,” says director of project management Ben Schrom."
I've worked in the mapping space for more than 20 years and I can tell you one of the biggest annoyances with using third party mapping data is that you are beholden to their cadence and tooling for making updates to the map - you have virtually no control over the data. This is a big reason both Google and Apple built their own map.
The benefit of using OSM is that you are provided decent tooling and a relatively streamlined process for making updates. Map data quality is incredibly important, especially for navigation and it's really hard to manage because the map is always changing - it's not a static thing.
I was part of a startup whose entire point was opening up that control. Sadly we failed to gain traction. At least our idea is still around and we may pick it back up if we find the time and energy.
> Second, it is highly antagonistic to Google - who is probably still a partner to Lyft in many other aspects (advertising etc). Taking shots at your partners is generally considered bad for business.
Wasn't it Google that significantly raised their map api price a few years ago that prompted ride sharing apps to pursue open source solutions to replace google maps? Google took the first shot here. Other ride sharing apps are doing this. Grab is another ride sharing app that invest heavily in open street map.
It may be antagonistic, but Google Maps really isn't the one-stop shop it used to be in the mapping world. Companies like Lyft have been looking for alternatives for a while now.
The article I read talked a lot about the benefits to drivers. That translates fairly directly to the service being offered to customers. Having your driver more easily figure out where to pick you up or drop you off is a huge service win for customers. Keeping your driver more focussed on navigation translates not only to better trips but safer ones as well. The statement was that they did not want to rely on a 3rd party to generate 90% of the pixels your customers are looking at. How did you miss all this customer focus? I am not sure it is fair to say this was all about profit juicing for Lyft.
My issue with the article is just the mind-boggling premise that leveraging an Open Source mapping project is “the impossible” vs using Google Maps. Also, there are lots of other options including of course ESRI.
I have been in a company that marginally that was in the same situation as Lyft, financially speaking, and for me sounds brave and bonkers that those folks still have time to blog post in a very stock-nose-diving situation.
The title here is missing a key word from Lyft's post: "secret", for the "secret plan". I thought map ownership/dominance was a key idea, like a decade ago, when Lyft and Uber started. Interesting that the issue was OSM wasn't robust enough until now. Was Lyft actively assigning engineers to work on OSM (or waiting?); couldn't that have accelerated OSM's path to robustness? This idea certainly takes time to develop into reality, but it seems like a lackluster idea to post now, in hopes of making Lyft more robust itself, but it's hard to not be cynical about stock price bump PR releases.
Great can someone at Lyft either comment on if they will contribute what they made on top of OSM or if it is even useful to contribute back ? I think the answer is No and this is HN clickbait.
It’s weird that Lyft is flexing on something that Uber did almost 10 years ago. Everything they talk about has been known and done by Uber for a very long time.
> and pickup and drop-off guides to heavily trafficked destinations, so drivers and riders can get in and out of the airport (or stadium or park) more easily
Lyft doesn't have that?! I though this was a standard feature of taxi apps. Yandex Taxi allows you to call a taxi to a specific pole in the arrivals area at the airport in my city, for example.
> removing the distracting clutter of irrelevant stores, restaurants
This much at least makes strategic sense, with Google continuing to ratchet up intrusive advertising. How long I wonder before we see, say, animated icons for "selected" cell phone stores?
The way they made it sound seemed like they built a new solution for the maps. Them using OSM over Google Maps doesn't seem like it is too much of a big deal due to Lyft's limited presence.
In some sense OSM is "just" the database — what things exist and where are those things.
A mapping product like Google/Apple Maps includes a lot more functionality that often gets taken for granted.
For instance:
- A geocoding search engine to find places by name, address, or description
- navigational directions
- tiling and other data distribution (which is both software and hardware)
- styling and UI
There is a huge variety of open source options for this functionality, but it's hardly a plug and play thing, and quality varies wildly. Putting it all together is no small feat.
Disclaimer: I work on Headway (https://about.maps.earth) which is trying to make this kind of thing easier.
"For instance, Google always presents drivers with the fastest route, regardless of tolls. But sometimes, paying a toll would shave only a minute or two off the ride, while adding significantly to the cost."
Ah... the joy of NYC tolls. Even with EZ pass, a ride to JFK from NJ is almost $28 just in toll roundtrip. $37 without EZ pass. $37!!!
That seems fair for crowding into one of the most densely populated areas in the country, especially since going to JFK instead of flying out of Newark means you’re taking up space and polluting NYC without generating any positive value for the city. Tolls directly benefit the city by partially helping to pay for the extremely expensive infrastructure needed to make that trip work, and indirectly by encouraging people to make travel plans with fewer downsides for everyone around.
JFK and Newark don't have a lot of overlapping routes that I use. They used to pre-9/11, but a lot changed since then. Big international airlines and routes that I take don't fly out of Newark.
Believe me, if I had the choice, I would.
Thankfully, domestic routes are still decent from Newark, though often more expensive than LGA or JFK.
This isn't just about the airport. I visit friends in Queens and spend money there. The toll is exactly the same as going to JFK.
That's a curious use of an already established term, "Digital Twin"[0]. A digital twin is a virtual model designed to accurately reflect a physical object.
Link me? I'm not sure what you're referencing. For what it's worth, the concept and model of the digital twin that I referenced was first publicly introduced in 2002 by Michael Grieves, at a Society of Manufacturing Engineers conference in Troy, Michigan.
Facebook uses OpenStreetMaps and I'm pretty sure they had to design their own batch processes to validate that area's of interest weren't renamed racial slurs or for Pokimon creatures.
That was an issue because of the validation pipeline Mapbox was using at the time, the vandalism had long been fixed in OpenStreetMap proper when the story hit the news.
So, yes, you are going to need to make a trade off between potentially more incidents, but faster repair, and less incidents, but being stuck with your current validated release for a while if something goes wrong.
PS: it is OpenStreetMap, no plural "s"
PPS: Meta produces a publicly available validated version of OSM data.
They likely do use mapbox for rendering maps, as their SDK is really customizable, is designed to work with the OSM data format, and you don't need to pay mapbox to use it (at least for older versions). But mapbox is not necessarily a good solution for all of the routing requirements Lyft has. For example, at SFO ride hailing drivers are required to wait in a special parking lot outside of the airport, and then dispatching for rides is done based on which driver has been waiting in that lot the longest. You could probably make this work with mapbox, but it isn't necessarily a matter of just plugging it in and calling it done.
I'm curious to see how good Lyft Maps is, too. Uber Maps is, frankly, a disaster in my area, frequently suggesting illegal and/or impossible turns. I try to submit map feedback every trip (very hard as a rider, as most forms of available feedback seem to be to punish the driver rather than the platform), and it's still never been fixed. Most drivers seem to use Google Maps or Waze instead and if I see a driver using Uber Maps, I need to watch the trip and suggest several corrections. Ironically, I'm only a few miles from the office where most Uber Maps staff used to work.