Robot kills man



There are several ISO standards that relate to the contruction of robot and robot systems, these are robust but these need to be supplemented by safe working practices. From my experience, the UK and mainland Europe have good adherence, rest of the world is not so good. I was once inside a robot cell in Asia when someone removed my 'tag' that was on display to show that I was inside. I remember reading this story several years ago about a young woman killed in the USA just before her wedding.

Hyundai, Kia Supplier Faces $2.5 Million Fine in Death of Worker
 
There was no malice involved though. This is not an artificial intelligence issue. It was just an example of crap controls on a dangerous bit of machinery.
Computers will make much better drivers than humans. Putting humans in charge of a couple of tons of metal isn’t really a great idea, there are some terrible drivers about.
That part is absolutely true. However, the ultimate default safety net will be to stop the vehicles. There will be regular gridlock when situations arise that the processors cannot resolve, whereas a human could in milliseconds.
 
Last edited:
so a fella was killed when looking at a machine that was known to be faulty, and we are surprised that the faulty machine pretty much did what we knew it would do… sounds like it’s the humans at fault here.


Absolute failure of safe systems of work, the machine should have been isolated prior to inspection. Same people claiming this as the end of days will cite health and safety gone mad everytime someone steps in to stop them loosing an appendage.
It sounds like this.

Although people are worried about AI at the moment, there is a very good chance that this was a standard robot without any intelligence.

Think of Big Trac: Move forward 3, turn left 90 degrees, move forward 2, etc

It likely had a few sensors to detect when the crate was near but that is about it. I've been experimenting with a wheeled robot I made and it is surrounded by range sensors. It will run independent on the floor for ages and I've told it to veer slightly to the left if it detects it might collide with an object it has just detected on the right. It doesn't learn, it doesn't make decisions, just follows a set program, if then else....
 
It sounds like this.

Although people are worried about AI at the moment, there is a very good chance that this was a standard robot without any intelligence.

Think of Big Trac: Move forward 3, turn left 90 degrees, move forward 2, etc

It likely had a few sensors to detect when the crate was near but that is about it. I've been experimenting with a wheeled robot I made and it is surrounded by range sensors. It will run independent on the floor for ages and I've told it to veer slightly to the left if it detects it might collide with an object it has just detected on the right. It doesn't learn, it doesn't make decisions, just follows a set program, if then else....
Exactly right, but it will be spun by newspapers (and the online equivalents) as "Fear of AI" to get the clicks, reads, views and hits.
 
It isn’t, if the correct safety protocols had been in place or if in place followed I’m confident the poor bloke wouldn’t be dead.
So how's that any different from millions of others where the correct safety protocols weren't followed?!
 
Exactly right, but it will be spun by newspapers (and the online equivalents) as "Fear of AI" to get the clicks, reads, views and hits.
Because some people don't understand what AI is and of course spreading fear of new robot overlords is nice click bait and that is all that counts.
 

Back
Top