Otto joins Uber
blog.ot.to
blog.ot.to
The line 'When we started Otto we committed to rethinking transportation' - what, three months ago?
I hadn't heard of these guys before, but a business designed to be quick flipped? Was this business ever not part of Uber? What kind of three month old startup has a team that size?
There doesn't seem to actually be even an MVP product there. How do businesses like this end up getting built and sold so quickly?
'Kalanick began courting Levandowski this spring, broaching the possibility of an acquisition during a series of 10-mile night walks from the Soma neighborhood where Uber is also headquartered to the Golden Gate Bridge. The two men would leave their offices separately—to avoid being seen by employees, the press, or competitors. They’d grab takeout food, then rendezvous near the city’s Ferry Building. Levandowski says he saw a union as a way to bring the company’s trucks to market faster.'
Which implies to me that the plan was for Uber to acquire Otto all along.
"Okay, Travis, but I don't want to just get hired on as another employee, I want to be cool like Kyle Vogt. Give me a few months to buy some trucks, rig them up with sensors, put splashy logos on them and get some press, and then you can acquire me using some of those Saudi gigabucks you picked up in your series F."
https://medium.com/@sacca/why-i-would-never-want-to-compete-...
This kind of thing is pretty common at Adobe and Cisco as well. There is a bit of a revolving door of executives and senior engineers leaving to start companies, intending from the beginning of being acquired by their former employer if their business pans out.
Many companies could not take the risk of running an innovative new business internally -- their board wouldn't let them, or their investors wouldn't. This kind of thing probably results in more research and innovation taking place that would have otherwise.
You prove you can build the prototype, de-risking the hirer's decision to pay you more than a normal hire. This reasoning presumes that the 'normal' hiring process and market resembles a lemon market. This is my IMHO. Bad hires drag down all participant's perceived value and compensation.
Of course, what's "hot" can change pretty fast, so you have to be lucky as well.
We were also thinking of branding as Otto, and had a hard time finding anything about a company named Otto -- stealth mode indeed.
MY GOD. The bbq was in 2002, and they kept the site up and running through 2015 before selling it. This is like a piece of history or something.
[1] https://www.otto.de/ [2] https://en.wikipedia.org/wiki/Otto_GmbH
https://github.com/square/otto
Disclaimer: I'm an employee of Clearpath, who are behind OTTO Motors.
The founder started a company called 510 systems which was sold to Google, and years later he left Google to create another company focused on self-driving trucks. They obviously have deep experience in the area, and Uber obviously wants that expertise.
EDIT: this would be a subcategory of the acquihire
(You know, just like that real estate developer is always scrupulously honest when negotiating for that last plot of land, they always say something like "This is the last piece I need to make my seven digit profit! How much did you say you wanted for it?", and would _never_ try to tell the owner of that plot "This is exactly the same house design that my alzheimer's affected mother has lived in all her life - it'd make her last few years _so_ much more comfortable. Would you accept $market_rate*0.8?"...)
The area being flipping companies to large corporate buyers?
This case might be a little different given the short time this team existed though.
Considering that, IMO, Valley companies severely overestimate that risk I think acquihires are a really expensive way to acquire people.
Uber acqui-hires ex-Googlers
They quote: " It’s no surprise that trucking companies see over 90% of their drivers quit every year in search of better opportunities. " That was in 2014, and it was due to (from the link they provide) “larger fleets raising pay, offering bonuses and attracting more and more drivers from smaller fleets to fill seats.” They were still truckers, just different companies.
The Otto piece made it sound like they were leaving trucking all together.
The whole tone of the article of "making lives better for truckers" just falls flat. Not sure why they're trying to suger-coat it.
If your job category is done there is no use fighting it. Just make sure your kids have another choice.
[1] https://en.wikipedia.org/wiki/UK_miners%27_strike_(1984–85)
Probably with gradually reduced overall pay, because there will be a fraction of the jobs with the same number of potential truckers. I'm sure investors will see a solid return though.
I love how SV thinks everyone wants and needs their "help."
Source: http://www.lefigaro.fr/conjoncture/2016/06/08/20002-20160608...
In Germany the public railway company Deutsche Bahn makes more than 1 billion in profits. All the profits come from freight, transporting people is done at a loss. The problem is that not enough money gets invested into upgrading lines highly frequented by freight.
It's fine for public transportation to not make back the investment if the advantages to the public are high enough (though with the Franc that's easier to deal with than the Euro)
Sounds like the broken window fallacy argument to me.
Now, freight and commercial trucks are effectively the biggest contributor to this tax plus they have to pay for weight. Consumers are the ones getting freebies (especially electric, hybrid etc.)
I couldn't disagree more! Wear on roads is to the fourth power of pressure. And while trucks might have 10x the contact patch your car does, they might weigh 30x as much. So the pressure might be 3x and thus the wear 3 x 3 x 3 x 3 = 81 times as much.
If you look at road costs as largely maintenance then cars are subsidizing trucks because one truck could do orders of magnitude more damage, but have to buy orders of magnitude more fuel. Why is that? Because past a certain point fuel consumption is dominated by drag, and thus, frontal area. 8-10 MPG for an 18-wheeler is reasonable. That's not 81 times worse fuel consumption, perhaps only 2x or 3x compared to modern cars.
The rule of thumb in civil engineering is actually fourth power of axle load, per axle.
> So the pressure might be 3x and thus the wear 3 x 3 x 3 x 3 = 81 times as much.
Considering the rule of thumb is by axle load (not pressure density) you're off by about two orders of magnitude when comparing a large sedan to an 18-wheeler:
* A Model S is 2T over 2 axles, or 2 * 1T/axle
* A fully loaded 18-wheeler can go up to 36T (legal limit) over 5 axles, or 5 * 7.2T/axle[0]
That means the road wear per mile of an 18-wheeler is up to 2.5 (number of axles) * 7.2 (axle load) ^ 4 (4th power), or 6700 times the road wear of a Model S.
Now that's a worst case scenario (unloaded model s versus fully loaded 18-wheeler), but you get the point.
[0] on average, the front axle is usually ~5.5T with the rear axles being in the 7.5~8T range which increases the road wear difference quite a bit, to about 7500
Trucks pay fuel tax on diesel fuel (can be higher rate than gasoline), surcharges on fuel tax in some states, weight-distance tax in some states, and heavy vehicle use tax.
The federal weight limit for a truck--including the weight of the truck and trailer--is 80,000 pounds (without special permits), but they rarely haul that much.
8-10 MPG for a truck under real-world conditions is extremely uncommon. Typical efficiency falls in the 5-8 range, depending on the type of cargo + weight, geographic area, etc.
This isn't quite right. The pressure on the outside of the tire is the same as the pressure on the inside, and truck tires aren't inflated to 3x the psi of car tires - they might be 55 psi instead of 35. If pressure were really the issue, bicycles would be much more damaging to the road. Racing bikes are typically inflated to 100-150 psi.
As your link points out, the correct relationship is linear in the fourth power of the weight per axle. Your conclusion is correct of course.
http://www.honestjohn.co.uk/forum/post/index.htm?t=79414
http://www.trucknetuk.com/phpBB/viewtopic.php?f=4&t=89338
(Doesn't change your assertion materially, of course. Your comment didn't scan correctly to me because I've seen video of arm injury caused when someone punctured a truck tyre with a knife. Meanwhile, I've deliberately punctured motorcycle tyres by hand at 36psi while performing a repair (boring out a hole for a plug).)
Nothing.
Well, the "pressure applied to the road" is approximately the same as the tire "inflation pressure"... It's not exactly the same because you need to take into account that the pressure in the tires increases (a bit) when you load the truck, and you have to be clear that you're talking about "overpressure", i.e. the relative pressure difference between the inside of the tire and the atmosphere.
What I think you meant to say is "What does the tyre inflation pressure have to do with the force applied to the road due to the gross vehicle mass?" Then your conclusion ("Nothing") is correct.
As another commenter pointed out, I've ridden my bike with 150psi tyre inflation pressure but the weight per wheel is my weight + bike weight devided by contact area of each tyre. Sure, higher pressure tyres typically have less contact area, but the gross weight of the bike is tiny compared to a truck the road wear from cycling is pretty much zero.
No, it's probably to the fourth power of force, not pressure. These are distinct and separate concepts in physics.
> We could assume the tyres were solid rubber for the purposes of this conversation and the wear on the road would be the same due to the gross vehicle mass devided by the number of tyres.
Ok, but in that case we must be talking about force, not pressure. Because the pressure is force divided by contact area...
> As another commenter pointed out, I've ridden my bike with 150psi tyre inflation pressure but the weight per wheel is my weight + bike weight devided by contact area of each tyre. Sure, higher pressure tyres typically have less contact area, but the gross weight of the bike is tiny compared to a truck the road wear from cycling is pretty much zero.
Now you're mixing pressure and force in a way that doesn't make any sense to me...
So dmurray is saying that "correct relationship is linear in the fourth power of the weight per axle" which IMHO is correct (but to fit with my argument above I would use the words "force exerted by tire on road" rather than "weight per axle", it's the same).
You (TheSpiceIsLife) are talking about something else... I don't know what... I can't say if you're right or wrong, but I feel quite certain that what you say is beside the point.
My (only) point was that the force that a wheel can exert on the road (what Newton called "action") is
A = [current "overpressure" inside tire] * [contact surface area]
R = [pressure on road] * [contact surface area]
EDIT: Extended to better explain context of my argument.
They're right. Civil engineering estimates of road wear for wheeled vehicles[0] is the fourth power of axle load (the weight borne by the axle) per axle, tire pressure does not enter the equation.
[0] following extensive testing in the 60s, repeated a few decades later, the exact exponent is variable but 4 has proven pretty good for a rule of thumb.
Ehuu... So why didn't you/they just upvote dmurray who said that 10 hours ago...?
Once a vehicle is loaded the air pressure of the tires can go up substantially relative to what they're inflated to. Like you said, action and reaction.
That's why under-inflated tires will have a very large contact patch; increasing the contact patch increases the area in the area * pressure equation. But it also reduces the volume of air inside the tire/tube and that raises the pressure, also.
Absolutely, and Newton surely didn't say so. :P
(A more close to home example is that the pressure inside a tire surely doesn't have to be the same as the pressure outside the tire. In fact, if it is then we call that a "flat tire". :P)
But the force that the tire exerts on the road (=the "action" as Newton called it) has to be equal to the force that the road exerts on the tire (=the "reaction" as Newton called it). Since the contact surface between tire and road is the same that means that the "overpressure" in the tire (=the pressure the tire exerts on the road) has to be the same as the pressure that the road exerts on the tire.
https://cdn.theatlantic.com/assets/media/img/posts/2015/05/r...
And of course that doesn't take the environmental externalities into account.
That makes sense, since the road wear increases superlinearly with the vehicle weight - so commercial trucks are also the one generating most road maintenance costs.
They contribute more to the tax because they are bigger, more powerful, potentially less efficient.
They have to pay for weight because the increased weight has a direct correlation in actively shortening the life of the road, above and beyond "vehicle-miles covered".
Except the number of miles driven (or where you drive them) has very little relation to how much damage you do to the road, and thus how much of the maintenance costs you generate.
Not really the case any more, I just renewed my tabs on our EV and there's now a $150 EV tax per year in WA state.
The eventual beneficiary of cheap transportation is the consumer. I suppose this should considered a regressive tax.
Amazingly, this is not categorically true:
https://en.m.wikipedia.org/wiki/Great_Northern_Railway_(U.S....
On the one hand, rail is much cheaper per-mile than traditional roads, for both construction and maintenance.
Rail is also more efficient in terms of goods hauled per unit of fuel spent, and controlling carbon emissions from rail is massively easier than doing so from individual trucks.
However, rail itself is fundamentally... unegalitarian, maybe? Rail lines can only be used by those with the capital to run rolling stock between well-specified endpoints (typically large urban centers), and on specified timetables.
The system as it exists in the US right now is less efficient, but opens up a lot of possibilities. For example, I live in Japan and don't own a car right now, which means that I basically can't go backpacking. But if I rent a car, then I can go out to more remote places that are less frequented by the non-hiking masses, and actually enjoy some of the sheer beauty of the Japanese countryside.
Self-driving cars and trucks really do seem to bridge the gap between "cheap rail transport" and "private vehicles for everything", but I worry that they will continue just to be a convenience for the well-off (which is what Uber and Lyft currently are, frankly), rather than a means by which to make transportation available to everybody.
A trucker's job would be to ride out to the edge of town and back again. Life would be like that of a UPS driver (though driving a big truck is real skill so it might command better pay). No more weeks away from home.
Too cumbersome to switch cars probably for an end user..
"BEEP! BEEP! Wake up Jim! Our simulation predicts we'll be getting into an accident after we turn the upcoming bend."
This seems highly unrealistic. If the vehicle can run a simulation and know that it's going to get into an accident it can slow down, stop, pull over or simply avoid the accident all together.
No if there is an emergency it's going to be an "oh shit" moment and the trucker is going to have to grab the wheel and try to correct for a situation they're already in.
but, there is something to be said about our calling it "unrealistic" and not "impossible". :)
http://blog.algorithmia.com/machine-learning-trends-future-a...
Sure there will be more hubs than there are ports, but it doesn't look good for truckers.
Not saying you were arguing any different op, just furthering the points made above and below.
It's hard to imagine that not a lot fewer unless you expect that number of trucks on roads would rise by some multiple of now.
It's true in the same way that we can say that "factory workers" in the US get paid much more than they used to. This is because they are more likely to be technicians now, who service the robotic assembly line. The nature of the job has changed with automation, and truck driving will be similar.
The last-mile drivers aren't debugging robot trucks. In the worst case the robot truck just turns back into a regular truck. No new skills required for a trucker.
But obviously I am not smart enough to back it up with anything or I would probably be getting acquired by Uber too.
But hardly enough to keep all current drivers employed.
By how much varies (https://en.wikipedia.org/wiki/Elasticity_(economics)) but not that it happens.
I'm talking about more "truck miles" being produced, while you're talking about the absolute number of trucks.
Now that we understand each other, there isn't much left to argue about :)
Then you could have a small number of vehicles, or small companies could pop up to serve this role, that refuel automated trucks that run out and stop driving outside those automated routes (if this is even a concern).
If you are eliminating 80%+ of trucker salaries, you have a lot of money to work with.
Or do you think they will benefit from resisting all forms of automation? I doubt it will slow down the eventual outcome, which is pretty much set in stone already. But maybe I'm wrong.
I'm all for this kind of technology but they need to at least be honest about what they're working towards.
After 2 years the original Otto product failed to get a footing in the market. Simultaneously, all attempts to integrate Otto technology into the core uber product failed due to internal politics and the fundamental impedance mismatch between self-driving AI for trucks and urban transportation. Additionally all original core Otto staff have left, and some are pursuing a new company that will revolutionize the future of truck driving.
Based on the timestamps, kinda looks like the Otto blogpost was prompted by the Bloomberg article -- even though the "agreements were reached" in July according to Bloomberg.
1 - https://news.ycombinator.com/item?id=12311559 2 - http://www.bloomberg.com/news/features/2016-08-18/uber-s-fir...
[1] http://spectrum.ieee.org/automaton/robotics/industrial-robot...
Nobody really wants to do these jobs, except maybe a bike messenger, and that only last so long.
Going from zero to self-driving seems a little advantageous.
Looks like they got beat to the punch big time here.
LIDAR is the only way to go, imo.