Why aren't the Japanese leading the way in self driving instead of Israel and U.S.?

calisnow said:

It just occurred to me that the Japanese have had a long standing cultural love of robotics. They are also very concerned with their aging population.

And if anyone recalls it was the Japanese who launched some of the initial major break-throughs in driving assistance / automation:

1996 - Acura launched the first in-car vehicle navigation system in the Legend.
2004 - Infiniti launched the first active cruise control in the QX56 SUV (I have that SUV - it's radar based and was breathtakingly cool at the time, despite how primitive it is and how clumsy and error-prone its implementation).
2013 - Infiniti launches the first hands-off lane-keeping system in the QX50 sedan. But it's primitive, severely limited in its capability and according to owners ping-pongs back and forth in lanes.

And that - is it for Japan - as far as I can remember.

Where is Honda? Toyota? Nissan?

You would think that highway autopilot would have launched on Japanese streets long before America.

Instead we have a couple of visionary lone genius types from Israel and South Africa (Musk, and the computer scientists at Mobileye) actually pushing the first functional self-driving cars on the road - leading the revolution.

All of which is fine - this is just curiosity on my part.

Click to expand...

the Japanese companies are leading in development if self driving cars, but the route to rmarket roll-out is different.

For similar reason why Hong Kong banned Tesla auto-pilot, Japanese companies will go slower to market.

It must be suitable for all members of society, particularly the old and frail.
Why it does and does not do must not be mangled in translation.

the result is that Tesla can deploy an beta grade deployment and use it gain end user, billion on road miles. A considerable competitive advantage vs the rest of the industry. But Toyota and Nissan will want a more robust system with more expensive sensors. The need for independance and dignity for the elderly demands that Japan will go robot car in a big way. But zero tolerance for a system that be 100% reliable 100% of the time (or more precisly, despite whatever a system is capable of doing, it will only released to do what it can do 100% reliable, 100% of the time)

calisnow said:

Where is Honda? Toyota? Nissan?

Click to expand...

Nissan says 2018 for the Leaf:

2018 Nissan LEAF will have 200 mile range, self-driving tech

US has a software advantage, no? Seems like autonomous needs really good software engineering. I'm not aware of Japan leading the world in there, say, compared with manufacturing and electronics.

tinm said:

US has a software advantage, no? Seems like autonomous needs really good software engineering. I'm not aware of Japan leading the world in there, say, compared with manufacturing and electronics.

Click to expand...

Japan is a powerhouse in software that does not rely on interfacing with the user primarily via words or numbers.

Ie Japan is a master at industrial and gaming software, but minnow at corporate and enduser software

renim said:

Japan is a powerhouse in software that does not rely on interfacing with the user primarily via words or numbers.
Ie Japan is a master at industrial and gaming software, but minnow at corporate and enduser software

Click to expand...

It's also hard to discount the vastly different environments. As a whole, Japan in far more urban than the US. It's unfortunate for them, then, that urban environments are more difficult for a self-driving car to navigate. They have a different primary use case then we do.

I went to Toyota Automobile Museum in Tokyo around 10 years ago. There was a self driving exhibition, there were dots on the closed road, the cars automatically steer to follow the dots, it was kind of fun.

calisnow said:

Where is Honda? Toyota? Nissan?.

Click to expand...

Nissan?

Nissan is a leader in autonomous car R&D. They have a clear roadmap until 2020:

Traffic jam pilot. By the end of 2016, Nissan will offer cars that are driving autonomously and safely on congested highways.
Fully automated parking systems. Also by the end of 2016, Nissan will offer fully automated parking systems across a wide range of vehicles.
Multiple lane controls. By 2018, the company plans to offer cars that autonomously negotiate hazards and change lanes.
Intersection autonomy. By 2020, Nissan will introduce intersection-autonomy, enabling vehicles to negotiate city cross-roads without driver intervention.

Mr. Ghosn was careful to draw a distinction between Nissan’s vision of Autonomous Drive vehicles and what he described as self-driving cars.
“I want to clarify that there is a big difference between autonomous drive technology championed by Nissan, and self-driving cars. Autonomous Drive is about relieving motorists of everyday tasks, particularly in congested or long-distance situations. The driver remains in control, at the wheel, of a car that is capable of doing more things automatically,” Mr. Ghosn said. “Self-driving cars, by comparison, don’t require any human intervention – and remain a long-way from commercial reality. They are suitable only for tightly-controlled road-environments, at slow speeds, and face a regulatory minefield.”

Click to expand...

Source: Nissan Lays Out Road Map for Autonomous Cars - The CIO Report - WSJ


On-Board Nissan's Self-Driving Car of the Future - Bloomberg Business (This is an earlier demo)

What it's like riding in a million-dollar autonomous Nissan Leaf - CNET (October 2015 demo)

Just as with EVs (Nissan-Renault is the clear global sales leader in EVs, many people don't know because there are more reports about Tesla and others...), their advances in ADAS and future robot cars often are under-reported by media outside of Asia.

They just presented the Nissan IDS Concept (hinting at both ADAS features and the next-gen Nissan Leaf):

Nissan IDS Concept: Nissans vision for the future of EVs and autonomous driving - Nissan Online Newsroom

renim said:

Japan is a powerhouse in software that does not rely on interfacing with the user primarily via words or numbers.

Ie Japan is a master at industrial and gaming software, but minnow at corporate and enduser software

Click to expand...

Culturally Japan is very good at engineering and making something that already exists better, but the way their R&D is structured, lower level people aren't given anywhere near the freedom of movement you see in American companies. It is expected lower level people will do the will of the boss without question. In US companies and especially US tech companies, lower level employees are often empowered to question anything that doesn't seem to work and introduce new ideas. Some US companies are structured to maximize this. This is a major advantage when it comes to software and firmware development.

Software is a mix of engineering and art. People who can walk both sides of those disciplines can be very good at software development. An organization that encourages communication without hierarchies allows the more artistic and more engineering oriented to cross pollinate their ideas and produce a better product.

This hierarchical thinking is not unique to Japan. I have a friend who is a software program manager and she's managed both US based teams and teams in India. She said Americans are much more proactive. If she didn't give her team direction every day, they would figure out for themselves what needed to be done next and usually got it right. She said the Indians would just sit there and do nothing if she wasn't constantly, actively managing them. She found the Indians a lot harder to work with. The Indians had been trained from an early age to follow authority and not step outside their bounds. The occasional rebel who broke free of the rigid thinking usually emigrated.

I also came across an article why China has so few nobel prizes despite having an education system that tops many international tests. Again, China's education system is focused on learning by rote and following authority instead of thinking for themselves.

My 2 cents...

Does @tftf sounds like Bertel Schmitt? Really this post for some reason made that connection for me...

(Bertel Schmitt - the guy who said he would eat a Tesla hat if Autopilot could do what it does within a year of announcement... then backed out of it after autopilot kicked ass... see here Autopilot a fraud? The linked author bets Elon he'll eat a hat if it changes lanes! )

The Japanese culture is methodically conservative and tends to not be terribly innovative. That is not a criticism at all. It serves them exceptionally well in product development. Look at their delivered product. Usually excellent.

We Americans tend to be the wild innovators. Sometimes with abandon. But we create really cool stuff that the world then follows. In the case of autonomous driving cars, based on just the small sample of idiot Tesla owners doing dangerous things with new tech, the Japanese are better served waiting for the largest car market in the world (for now) to make all the mistakes possible before delving into this exceptionally complex area.

My criticism about what Tesla has released with AP is that it is not the same as releasing a new version of iOS, Facebook, etc, where it is easy to fix without putting lives at risk. That is the software developing mentality: fix it when we see what is really not working. I am quite sure that no Japanese mfr would ever release something like AP without proving, to the best of their ability, that it would pose no danger to the driver or others on the road. Though I do applaud Tesla for pushing the envelope. I truly do, as they will now spur others to do even better.

While I love the innovation coming out of Tesla, and their creativity and boldness, I think that the user base is simply not ready to responsibly use that innovation, as has been proven by the ridiculous videos posted, which are ruining it for the rest of us.

As a boss of mine said many years ago: never underestimate the ignorance of the "great unwashed". We will very soon see the impact of their irresponsible actions.

calisnow said:

2004 - Infiniti launched the first active cruise control in the QX56 SUV (I have that SUV - it's radar based and was breathtakingly cool at the time, despite how primitive it is and how clumsy and error-prone its implementation).

Click to expand...

My 2002 Infiniti QX4 had intelligent radar-based cruise control.

- - - Updated - - -

tinm said:

US has a software advantage, no? Seems like autonomous needs really good software engineering. I'm not aware of Japan leading the world in there, say, compared with manufacturing and electronics.

Click to expand...

This is a good point. A 360 degree camera system with forward radar and ultrasonics could do full autonomy with enough processing power and the right software. Once you can see around the entire car visually with enough resolution and a high enough frame rate, it's suddenly possible to drive better than any human.

The sensor hardware will be a commodity. It's taking that data and making it actionable that will have value.

Subaru has done really good with their Eyesight cameras, top rated front collision avoidance of any company (I had a Subaru Legacy before my Tesla and it prevented multiple accidents for me). They need to work on autosteer though, they finally added lane keep assist in the 2016 models.

CHG-ON said:

The Japanese culture is methodically conservative and tends to not be terribly innovative. That is not a criticism at all. It serves them exceptionally well in product development. Look at their delivered product. Usually excellent.

We Americans tend to be the wild innovators. Sometimes with abandon. But we create really cool stuff that the world then follows. In the case of autonomous driving cars, based on just the small sample of idiot Tesla owners doing dangerous things with new tech, the Japanese are better served waiting for the largest car market in the world (for now) to make all the mistakes possible before delving into this exceptionally complex area.

My criticism about what Tesla has released with AP is that it is not the same as releasing a new version of iOS, Facebook, etc, where it is easy to fix without putting lives at risk. That is the software developing mentality: fix it when we see what is really not working. I am quite sure that no Japanese mfr would ever release something like AP without proving, to the best of their ability, that it would pose no danger to the driver or others on the road. Though I do applaud Tesla for pushing the envelope. I truly do, as they will now spur others to do even better.

While I love the innovation coming out of Tesla, and their creativity and boldness, I think that the user base is simply not ready to responsibly use that innovation, as has been proven by the ridiculous videos posted, which are ruining it for the rest of us.

As a boss of mine said many years ago: never underestimate the ignorance of the "great unwashed". We will very soon see the impact of their irresponsible actions.

Click to expand...

I see both sides of the issue. On the one hand auto pilot really is just a next generation cruise control and the driver should be in control at all times. However, people will do stupid things with new technology. When someone is using auto pilot irresponsibly and hits someone, Tesla will be included in the lawsuit, but because the driver accepted responsibility before enabling auto pilot and they were told the limitations and responsibilities, it will probably fall back on the driver and Tesla will get out of the lawsuit unless there is some flaw in the system Tesla did not anticipate that made things worse.

Back when cruise control was fairly new I recall my father telling me about some guy who came to the US and rented a motorhome with cruise control. The idiot thought it was auto driving and he went into the back to get something and the thing went off the road. He was responsible because he was not using the vehicle as it was supposed to be used.

It is possible some sort of defect will be found out in the field that nobody anticipated and Tesla would be more liable for that. However it took a year to be delivered in large part because they tested the heck out of it under a wide range of conditions over the last year. If Tesla was smart, they would have brought in some of the FAA guys who dream up disaster scenarios for planes and make aircraft manufacturers jump through hoops proving the probability of the plane going down because of it is below a certain threshold.

Some more info and reviews on Nissan's pilot system (see my earlier comment for basic links) below:

Green Car Congress: Nissan begins testing piloted driving on both city roads and highway

Nissan Leaf Autonomous Drive first ride review | Auto Express