Why do Uber and Lyft have to develop their own self driving tech? If what they sell is their routing algorithm for drivers, they can simply lease self driving cars from the companies directly and run them on their network. The car companies would be happy to get a cut of the profits.
What advantage would they get by developing their own tech? They have to manufacture the cars themselves or lease the tech to other companies to cover the costs.
Uber and more so Lyft seem late to the game and Uber in particular seem to be fraught with problems and have not made the progress they hoped. So why pursue that? I understand if they started early and are making great progress. On the flip side is the tech no where near completion? Then that would make sense. But Google seems to indicate that they are very very close to launching their own taxi program.
They haven't succeeded in their initial goal of exterminating the existing cab companies, and that leaves them in a situation where developing self-driving technology could be a matter of life and death for them. The incumbent cab companies are much better positioned to profit from self-driving cars. They've already got all the infrastructure, expertise and staff it takes to manage fleets of vehicles. The bigger ones already have their own apps (e.g., Curb). If self-driving cars hit the general market first, basically all they have to do to run Uber and Lift out of town is buy some and factor the fact that they aren't paying cab drivers anymore into their prices.
By contrast, most of Uber and Lyft's basic business models right now are predicated on the idea of hiring humans, who in turn manage their own cars. That has deep implications: It means they have zero expertise in fleet management. They don't have any of their own garages for storing cars and handling maintenance. They do have apps, but the most valuable parts of their apps are the parts for managing humans. Incentive mechanisms designed to try and ensure that supply of rides and demand for rides remain relatively stable, for example. Their existing business operations are heavily built around recruiting humans - both riders and drivers. None of that is particularly useful in a world with self-driving cars, and the bits that are - basically just the mechanics of enabling people to hail rides and pay for them - are drastically easy to copy.
That said, worth noting that the fleet management/maintenance expertise will be of unknown/unproven value for managing self-driving cars. No domain experience; just spitballing...
- Unmanned SDC might have significantly different refuel/recharge/cleaning/roadside assistance/accident-response needs. - Unmanned SDC availability/hours/mileage patterns are likely to differ significantly over time when they aren't tethered to bodily functions. - The size and location of fixed fleet-management infrastructure may be suboptimal for different patterns; they might start off as an advantage but become a drag as players with less inertia optimize. - SDC sensor maintenance is going to be a big unknown to these companies. - Their liquidity and vehicle retirement/acquisition processes may not be compatible with rapid fleet replacement without outside help. - Even if they do manage a rapid replacement cycle, the throughput of their fleet management processes may be closely tied to their normal acquisition cycles. - Unless they're going to choose that moment to transition to an app-only-ride-request model, they may have to overcome significant design/software/hardware dev challenges of their own to handle non-app customers (i.e., being hailed on the street, taking verbal directions from people who may be drunk or not know where they're going, stop requests, destination changes...).
Nobody really knows what company will develop truly practical SDC tech first. Therefore nobody really knows what revenue model that company will have.
If a vehicle manufacturer does it first, then probably their main goal is still to sell equipment, so they would probably be open to selling/leasing it to anyone and everyone.
If a tech company gets there first (or any company without an existing revenue stream), there's always a chance this company will want it all for themselves.
Sure, Uber and Lyft have the name recognition. And they have the active users (accounts created, apps installed, payment methods on file). But SDC is such a killer tech that whoever has it might decide it outweighs everything else and they can build their own app and get the users to come over.
If you're Uber/Lyft and this happens, you could be completely left out. Not only is your service less sexy, you're also paying labor costs that your SDC competitor isn't.
I'm skeptical. This involves a tremendous capital outlay. Even if you solve that problem, you have to get the production lines running, logistical networks humming, and marketing gears going. Much simpler to buy or partner with one of the companies who have spent years training consumers to press a button and unquestioningly get into a car.
No, bu the ones with all the patents will be Waymo, Uber, GM.. Lyft?
How does this work? Seems like Waymo would have been able to pick up enough of the obvious ones to kill competition for 20 years. Or just MAD between the big players?
Same with every company in the current climate. Develop every damn thing yourself for fear of anyone having financial power over you.
It really is tiring. Yet, I can't blame them too much. This is where I wish the government would step it. What do I care which company has the best self driving AI? Ffs, define best? It's likely the safest. So we're literally letting them compete for safety - knowing that some will be less safe than others.
There should be a government program here to share knowledge, license startups and ensure equal progression for the benefit of public safety.
I really hate that our safety is being handled by Uber of all companies. Especially since they've already killed someone.
It may be the case of just throw enough ideas together and that works, in which case it may be effectively a commodity; everybody working on it has enough patents that cross-licensing works. Or it might be there's one key idea that makes everything work, and whoever files first controls it for whatever a patent lasts these days. Kind of depends which company owns that patent how it goes.
I doubt Uber wants to sit around and let Lyft develop self-driving tech, since I doubt Lyft would lease that tech to Uber on very favorable terms.
They likely expect the first company to come up with true self driving cars to start their own competitive service to Uber+Lyft. To them, it's an existential threat to not invest in their own self driving technology.
I think every self-driving car project is under the impression that the first stable and safe self driving car network is going to have no incentive to lease the technology out.
I think it's likely that if Uber-or-Lyft don't get a sustainable technological advantage with driverless cars, they will actually stick around and make a deal with the company that provides the driverless cars, but they will have a weak negotiating position and will have to accept a small percentage of the profit which will not justify their current valuations (especially Uber, which has a much higher valuation than Lyft).
Another instance of the "more data is better" fallacy. Humans can drive cars safely after only fifteen years of intermittent sensory input. Lyft could collect that "data" within just one year employing ten collectors. That still doesn't give you brains.
Is Techcrunch an outlet for PR pieces? I'm asking because that article reads like one of those.
More data is better. The fact that humans have better algorithms that need less data does absolutely nothing to negate this.
There are machine learning techniques that generalise well from few data, but they are not very well known in the industry.
Also, though more speculatively, I think the idea of "lots of data" is attractive to marketing departments. There's something about algorithms that need huge amounts of data and compute, that only a select few companies can use. I guess it gives bragging rights, of a sort: "we got the biggest data around. Buy our stuff!".
But like I say, that's speculation on my part.
It would likely be more informative to instrument a few cars with some advanced sensor package and let well ranked drivers drive them around than to try to gather data from smartphones in existing cars, but I suppose it depends on what the end use is.
Very narrow view point for you to take.
Using uncalibrated and random positioned phone cameras to learn about driving a car that has better sensory equipment seems backwards to me. But point taken, the article says "learn more about driving overall." So that could be anything.
Why are we still talking about Level 5 autonomous driving when we can't even get Level 4 working properly? I believe this is sending the wrong message.
On the topic of the acquisition, it seems like a good strategic buy for Lyft. I am still not sure whether they have the capital nor the talent pool to develop a strong autonomous vehicle product. It also seems like they are a bit late to the party, as there are more and more doubts on the reality of self-driving in the next few years.
I may sound sceptical, but I am just cautious when it comes to news on self-driving cars. However I am genuinely excited at this development and look forward to hearing more about how Lyft is progressing on this quest.
i think it is capability gap between tech companies and car companies. Car companies can't even get cameras around all the body to avoid accidental scratches during parking. Where is tech companies, while could easily do a lot of car tech, have no business case doing anything less than Level 5 - the Level 5 is a tech platform where is anything less is an advanced car, and the tech companies are in the platform business, not car business.
There is a lot of business to be had with Level 4. While the car may not be able to go everywhere, in restricted areas a robot-taxi/shuttle service would be a great advancement. We would all love to be at Level 5 but I would take a solid Level 4 in a few years instead of waiting much much longer for a viable Level 5 solution.
This has been a feature introduced by most car companies in the last ten or so years.
By the way, I don't think any Tesla car has this.
If Waymo's cars don't have usability issues and launch this year, Lyft's valuation will plummet.
