Automated train travels in wrong direction, injuring 20 in Yokohama

I don't like automated trains, planes, or automobiles. I'd like to see all of these manned in the future.

We are going to see more and more of these types of accidents as we move into the future of having automated vehicles! We already have airplanes that fly on auto-pilot, but thankfully there are pilots in the seats controlling the most important functions, and I believe that trains and other forms of transportation should also have a "human" there to ensure accidents like these do not happen!

How many lives must be taken before they "perfect the AI"??? Enough already with this AI nonsense.

We also have accidents when humans are in charge. Should we expect anything different from automated trains?

If it’s not my foot that covers the brake then I don’t feel safe...

Road accidents happen everyday. AI vehicles accidents barely happen.

How many lives must be taken before they "perfect the AI"??? Enough already with this AI nonsense.

There is a problem with using the acronym "AI" here and in other places as well. Typically people are thinking that "AI" is "artificial intelligence" when in fact all AI means here is "Advanced Intelligence" which is incorrect as well as it SHOULD be referred to as "AA" or "Advanced Analytics"

True "AI" will probably come sometime down the road, and will probably be widely used as well, but until then, we should be aware of the differences!

Advanced Analytics is a term that refers to a wide range of analytics tools and techniques which includes data mining, machine learning, forecasting, and pattern matching. Advanced analytics is also a marketing term used to categorize analytics capabilities beyond basic business intelligence (BI) solutions. Advanced analytics covers a wide range of analytics including predictive, prescriptive, and AI-driven.

Artificial intelligence is a set of technologies that simulate human intelligence. AI technologies aim to mimic the human ability to analyze and draw conclusions from data, understand complex concepts, and interact with humans in a human-like way. AI technologies are also trained to process and understand natural language and become self-learning.

https://www.zylotech.com/blog/advanced-analytics-vs.-artificial-intelligence-how-are-they-different

Who said that AI was involved? This line opened in 1989, long before AI existed. An automated train moving along on a guideway does NOT use AI. This accident is probably mechanical or electrical, based on sensors/motors/gears etc. Cars that use autopilot functions to make complex 'decisions' about braking and lane changes (e.g. Tesla) us AI to evaluate changing highway situations and respond. The automated SkyTrain system in Vancouver has operated safely for more than 30 years without problems, and is a far more complex system (53 stations/80km) than this short Yokohama line (14 stations/11 km).

I've ridden the Yokohama Seaside line many times going to uminokoen and Hakkeijima Sea Paradise. This is the first time I've heard of any trouble with it

And calling them trains is incorrect. They run on rubber truck tires like trams you might see at an amusement park

@kokuzi, that's also AI for sure but not like now, you can say analog level of AI not 5G

Anyway, for all cases, machine can't be replacement of human.

I've been on it several times. This line has been around for a loooong time. It's not as if it started operations a few weeks ago.

It’s likeky the accident was caused by human error, either in the control room or by faulty maintenance. This is an older system and should not fail.

Yokohama has been operating its automated trains for many years. This is unusual to hear and raises questions about its software. For instance, weren’t there fail-safes in the software to avoid crashing into the bumper post? My concern is that the software was not robust enough to handle this type of situation. Having worked with many engineers in Japan, I was often shocked by assumptions Japanese software engineers would make, such as assuming a system should never get confused, therefore, they felt no need to program for such edge cases. Time to take a second look at the software and plan for the unexpected.

People are safe that's a good news.Sometime it happens.Lets not see tje negative view only. Seaside line had been providing good services for many years.

"Skynet has revolted against humanity."

in fact... A bracket or something to mark the track side for correcting the position was removed by mistake. Or it may be in the state where the sensor by the side of the vehicle can not be detected by wind and rain for many years. Things happen.

Remember the metro in Lausanne, Switzerland. Totally automated. It ran both directions and part of the tracks were single sided. Never heard of an accident.

That being said, the technology is there since years ago. The man behind the keyboard is the problem, ironically

this is just symptomatic of the serious decline in quality in Japan.

From cars to appliances, those days of Japanese quality are long gone.

I am happy that nobody was killed. One accident is not going to make me boycott these trains like I don't boycott cars and planes.

We also have accidents when humans are in charge. Should we expect anything different from automated trains?

Yes. We should. That's the whole point of them. Or, at least that's how it's marketed.

But, most of us can see through the BS, and know that the whole point of automated trains, and driverless cars, etc are to cut costs for corporations who don't want to pay people because they view people as an expense rather than the asset they really are.

We are going to see more and more of these types of accidents as we move into the future of having automated vehicles!

of course, as more things become automated more accidents will occur, but when you compare it to the rate of human accidents and deaths people will come to realize that automation will be safer than human error. automation is inevitable.

Yes. We should. That's the whole point of them. Or, at least that's how it's marketed.

But, most of us can see through the BS, and know that the whole point of automated trains, and driverless cars, etc are to cut costs for corporations who don't want to pay people because they view people as an expense rather than the asset they really are.

More like most people are paranoid of anything new and innovating. You know why we continue to see these types of accidents? Naysayers like you who would rather go backwards than forwards keeping things from becoming something better.

To say that this is somehow connected to the quality of the design and manufacture of a train system built 30 years ago, in comparison to today's higher standards and somehow equating that with the malfunction, is a joke.

This. I'm the first to ridicule Silicon Valley as it seeks to disrupt entire industries with driverless tech or even Amazon sending out fleets of drones like mosquitoes. But these kind of automated lines have been running all over the world for decades almost entirely without incident. See here:

https://en.wikipedia.org/wiki/Automated_guideway_transit

I frequently use one of Kobe's two lines, in operation since the 1980s. They're perfectly safe. I worry far more when riding the suicide-line (JR).

Burning Bush: "But we're gleefully supposed to accept AI driven cars as "safe""

You should maybe learn the definition of "AI".

Computers and automated systems are failure prone, just like humans.

Arturo: "Computers and automated systems are failure prone, just like humans."

Not at all the same. Not sure I'm surprised you think so, though.

Whatever the age and technology of the system, the story highlights the underlying and quite proper distrust of systems we do not understand or have any control over being forced on us by overbearing, arrogant zealot organisations driven purely by their own profit. Society at large needs to take back control of the implementation of technology and only allow what is in our best interests as a society and not just blindly grab at the latest shiny shiny sold to us at great profit. Just because you can do something doesn’t mean you should.

injured but not dead.

the manager in charge of the software should face charges.

automating 12-kms of line?

what about the yurikamome line?

the manager in charge of the software should face charges.

It's a railway that has operated safely for 30 years. Brie seeking to press charges you should understand what the cause was and whether the software manager was negligent.

On the whole these types of train are much much safer.

An automaton is not AI!. It is simply a robot programmed to do a simple task.

I am happy that nobody was killed. One accident is not going to make me boycott these trains like I don't boycott cars and planes.

Huh?

Computers and software still break down and crash often. Can't trust them too much.

Computers and software still break down and crash often. Can't trust them too much.

Automated systems are evidentially much safer than manually driven modes of transport.

Usual idiots out in force. There are a few AI accidents, so let's scrap AI in favour of thousands of human accidents, that is obviously better..

Oops, still has “bugs” in the operating system, me thinks it’s back to the drawing board !

If we went back to the drawing board any time a bug was found, there would literally be no computers, computer programs, or hardware. None.

Many good points being made about adopting immature technology - sure, these things should be tested well.

However, the aviation industry has made great strides in improving air travel safety through fixing and evolving design or other related features. This is usually at the cost of 200 +- human lives, as it's only after a big crash that we are able to investigate and determine what went wrong.

What I'm trying to get at is, 14 or 20 injured is a small cost to pay for what will hopefully be an improvement in this automated train technology.