The problem, as far as I can tell, is as follows: top management is able and motivated to increase the size of the money bag. Developers just want to do a good job for a good salary. Middle management can't change the size of the money bag too much, so it just wants to grab as much as possible for itself. How does a middle manager expand their empire? Certainly not by leading a small, effective organization. The head count needs to grow, so the middle manager wants ineffective developers. It only works because no one above knows enough about software to fairly evaluate the work of these petty empire-builders.
One time I wanted to gauge how far the car industry has come from realizing the problem to implementing a solution, so I typed something like "automotive software crisis" into Google. Among the first few hits a presentation of a German CS professor stating that there is a crisis... and that the solution is model-driven development. It's the exact shit that they are already doing that is not working.
If good software is produced in Germany, it's despite subpar academic CS and dysfunctional large companies. Much of Germany's economy is medium-sized companies - fortunately, I guess.
I do expect the German car industry to fix its software but only when it's really suffering financially. Car industry outsiders will be heavily involved in the solution.
Another part of the problem is that in many industries there's a strong lobby of local companies which like medieval guilds block any new competitors from entering the markets. APIs and open standards still barely exist - you cannot simply plug in your new fancy service into an enterprise environment, but you have to pay plenty of money to some Verband to get some crazy XML specification for file exchange. After all I've seen in IT in Eastern Europe (most of all in Russia) and USA, I'm feeling like I'm still in XX century.
Long story short they are forcing us to do things as they always been done even though they asked us to change them in the first place.
It is not. I worked in 3 people startups, 15 people web shops, 50 people IoT shops, and very very big enterprise software corps in Germany and I can tell you what you say can be copy&pasted into any of them. The only advantage of smaller companies is that they can't afford to have as much middle management. Therefore the "visionaries" are forced to live closer to reality and work with small teams. But they certainly don't like it and move to your model as soon as possible.
Most docs at my >200 dev shop including contractors are in German though.
https://www.bloomberg.com/opinion/articles/2019-03-02/the-en...
There is a lot of really good CS/math research coming from German uni's, I would not call it subpar.
I've seen this in every company I've worked in. Importance of projects is measured in how many people work on them; you'd think that creating efficiencies and managing the work in an extra small team would be valued, instead you fall into complete irrelevance.
Reinvention? What sort of? They don't get software, they don't get new business models. That's why they stick to politics who should save their models - which is protectionism.
Germany's car makers were never customer first. What kind of innovation did Germany establish besides ABS? The cars still drive forward, backwards and can turn around - AWESOME! This has been done for almost 100 years.
Germany's car makers are great at marketing. "Drive a M5 and feel the power of 500+ horsepower on the Autobahn" - that's exactly why they are struggling with the idea of autonomous driving: how can you reinvent the story from being an Autobahn race car driver to calmness in your car with millions of other people? Same goes for Porsche: "Reinventing the classic" It still has enough HP to accelerate you from 0 to 100 km/h in roughly 4,2 seconds - and then comes Tesla and beats the sh*t out of Porsche without even hard work.
Germany's car industry is like the National Rifle Association. They won't change without hard disruption. And in Germany it finally arrived.
Not sure why you got downvoted, afaik the most important technological breakthroughs when it came to cars in Europe were made by the Italians (Fiat, mostly, but Alfa Romeo too), the French (the Citroen DS is one of the best technological products ever) and the Swedes (Volvo and its emphasis on security). Somehow the Germans navigated the ‘90s and early 2000s a little better, not sure exactly why (maybe because they invested in China earlier?), and that’s why right now they’re seen as the car companies to beat.
Carl Benz - invented and patented the first "motorcar"
Gottfried Daimler - also independently built an automobile the same year as Benz.
Siegfried Marcus - 131 patents, first petrol-powered vehicle.
etc.
[1] https://en.m.wikipedia.org/wiki/Tatra_97#Resemblance_to_Volk...
You can be as innovative as you wish, but if you are cheap and provide low quality products, that's what you get. Eventually someone is able to provide the same technology and then quality wins.
Mercedes and BMW are still very nice to drive in. The issue is more that German's generally don't believe in fads, and self-driving cars are one of them.
I just don't want a self driving car, which takes away the last bit of fun of sitting in a metal box: Being in charge of where the damn box goes.
Germans just have a higher bar for "innovation": It has to be useful.
Yeah, markets are different. In the US, manual transmissions are almost unheard of these days, in Germany they have a market share of >70% overall and >90% in small cars.
Somehow they don't find credit cards useful and still pay a lot in cash, when the rest of the world already embraced wireless payments. :)
The rest of the world embraced wireless payments whilst Germans still have a realistic view of privacy impacts. Being at the tail end of that particular adoption curve may still be seen as a feature in Germany. :)
I wish their view were a little more widespread.
You can pay with credit card for practically anything in Germany. Besides that fact I personally really like to have the option to pay in cash. I don't want literally everything logged by my bank.
How about: the car (Carl Benz)? And the motorcycle (Daimler)?
All the common (and uncommon) engine types: Otto, Diesel, Wankel.
More recently, always-on four wheel drive (Quattro). I am sure I am leaving out a ton of innovation in engine and drivetrain technologies.
> Germany's car makers are great at marketing
That's a fairly new development. German car companies were (and still are) very much engineering led. See, for example, Ferdinand Porsche or more recently Piëch.
What you are describing is the US car industry, which has been absolutely marketing-dominated since at least the 50s.
In fact, I'd argue that the current problems are, if anything, a result of the engineering-centeredness, which has a hard time dealing with disruptive innovation. A key feature of a disruptive innovation is that it is demonstrably worse than the incumbent, so engineers will dismiss it as "not good enough" and "toy", and be right. They just miss that there are new criteria, at which the new innovation beats the pants off the old.
I am pretty sure that if you discuss the current situation with German automative engineers, they will be able to show to you very convincingly that electric cars are a joke, don't perform adequately and aren't even environmentally superior. And that electric cars and particularly self-driving cars are marketing gimmicks.
But customer sentiment has shifted (marketing side) and technology and engineering will follow.
I’d be very curious to hear how the would argue for _that_.
Oh and their fuel, the electricity ? Mostly comes from coal power plants which, despite TRIPLE the efficientcy actually still managed to produce more CO2 ...
Electrical cars are heavier because of the batteries, so in practice it's not quite triple the efficiency.
Electrical cars are very environmentally friendly at the location they're used. On balance, they're probably about even or slightly worse than petrol cars.
There's a simplification many economists argue for. In our modern world, price is a proxy for energy expenditure, which is what mostly matters when it comes to pollution. Therefore cheapest thing (measured over it's lifetime of course) is going to be the most environmentally friendly. This price takes many more things into account than any analysis ever will and therefore be more accurate. By that measure, electrical cars definitely lose by around a factor of 2 at the moment.
>> ..German automative engineers, they will be able to show ..
Note that I personally don't necessarily agree, but the arguments are not weak.
Another point is that a good Diesel engine, and they have apparently fixed the emissions problems, is actually amazingly efficient, and on longer distances beats electric vehicles when you take into account the whole fuel to wheel chain for the current grid mix.
When you look at the greater weight (= production costs = environmental impact) of electric vehicles, that has to be amortised over the life of the vehicle. When the vehicle does mostly long distance, it is not going to be able to do that, because the Diesel is actually more efficient.
However, vehicles that primarily do short distances tend to actually not be driven much at all, they mostly sit around taking up parking space. So those aren't going to amortise the greater environmental investment either.
The exception would be vehicles that get driven a lot, but over short distances, so buses and taxis, but that's not a huge percentage of the total fleet. (And gives you a link between electric and self-driving, a link that otherwise doesn't make a lot of sense: converting to self-driving on-demand/rental cars is one of the ways to get electric to have a positive total impact).
I know engineers how left Opel after the merger with PSA because PSA is that what you describe. At Opel it was much different.
It used to be also different at Mercedes. Engineers are leaving right now Mercedes because they "streamline" their engineering. Which basically means cutting cost in terms that engineers who worked for the next generation cars now have to develop as well for the series production. THAT is killing innovation. Because developing for series production means developing with cutting costs in mind, like can we spare this resistor or that memory.
Al the engineers I worked with are not working for the series production but always thinking 10 years ahead and planing, developing and testing new ideas.
Apple killed its "automotive business" not without a reason. Because automotive is very much different than IT software. Automotive systems are highly connected embedded device. The next BMW 7 series released in 2021 will have 20 different CAN bus based networks. Automotive development means also development with preventing bugs in the first place. Because finding bugs by testing in these highly complex systems is next to impossible. This development style makes development slow. Because it is not the mind set "we can fix that later". Updates are expansive. Previously it meant they need to pay the service stations. Today with OTA it means to pay the telecoms.
Automotive development means also thinking about obsolesce. Yes, people get every 2 years a new phone. But what it with old once. An iPhone 4 these days is almost completely unusable. While some people have their cars new every 3 years, the average car usage is 16 years, in the US, IIRC. That means the average car is older than the smartphone as we know it. Automotive development means to develop with predicting the future 15 to 20 years ahead. Parts needs to be available, because people want to have their car serviced.
I just want to remind that Teslas with AP1 don't get no more updates. That is breaking his promise that all Teslas, all as in ALL, will improve by OTA. That is how old?
I know that other OEM are going the same route. But I don't see this as a disruption or innovation at all. It just means more land fill and is an environmental disaster.
BTW: My truck for pleasure is over 44 years old with no electronics at all. My car is a 18 year old Volvo, the first were Volvo introduced CAN and LIN, and I know the engineer who where responsible for that at that time.
This is simply incorrect. Their Autopilot software doesn't really get updated, but the majority of recent feature additions are probably not Autopilot-related. All of these changes are still provided to Teslas from 2012, for instance.