SoylentNews is people
SoylentNews
Submitted via IRC for takyon
Mining co. says first autonomous freight train network fully operational
On Friday, major mining corporation Rio Tinto announced that its AutoHaul autonomous train system in Western Australia had logged more than 1 million km (620,000 mi) since July 2018, S&P Global Platts reported. Rio Tinto calls its now-fully-operational autonomous train system the biggest robot in the world.
The train system serves 14 mines that deliver to four port terminals. Two mines that are closest to a port terminal will retain human engineers because they are very short lines, according to Perth Now.
The train system took ten years to build and cost Rio Tinto AUD $1.3 billion (USD $916 million) to implement. The trains are remotely monitored by a crew located 1,500 km (932 mi) away in Perth.
According to the mining company, the autonomous trains make sure the rails are clear ahead and monitor internal systems as well, checking for faulty wheels or couplers and bringing the train to a stop if there's a problem.
(Score: 1, Funny) by Anonymous Coward on Wednesday January 02 2019, @08:57AM (15 children)
Plenty of jobs for those Trump-voting coal miners when it get deployed in US.
(Score: 2) by MichaelDavidCrawford on Wednesday January 02 2019, @09:16AM (14 children)
Surely there is some other reason.
_Please_ tell me there is some other reason!
Yes I Have No Bananas. [gofundme.com]
(Score: 2) by MostCynical on Wednesday January 02 2019, @09:50AM (1 child)
theoretically, the autonomous trains are safer, and can run for longer (no need to stop to swap drivers), but they are automating everything [thewest.com.au]
"I guess once you start doubting, there's no end to it." -Batou, Ghost in the Shell: Stand Alone Complex
(Score: 2) by c0lo on Wednesday January 02 2019, @10:41AM
Even underground [northparkes.com] mines [itnews.com.au] of all sorts [abc.net.au]
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
(Score: 3, Informative) by zocalo on Wednesday January 02 2019, @09:51AM (3 children)
UNIX? They're not even circumcised! Savages!
(Score: 2) by Nuke on Wednesday January 02 2019, @11:29AM (2 children)
That's an ambiguous statement - I should think that figure covers the whole railway (I'm using the UK terminology), not just the automation, but the phrase "train system" impies just the automation. Trouble is that most journos (like most people) know next to nothing about railways. No-one who does would refer to a railway (ie the entire system, trains, track, signals, staff) as a "train system", they would call it a "railway", or "railroad" in US speak.
I don't undertand why it is supposed to be so hard and amazing to automate a railway. Zocalo cites the London Dockland Railway, but the London Underground Victoria line was automatic much earlier than that - 1967 - without even involving computer tech, and has run reliably ever since.
(Score: 0) by Anonymous Coward on Wednesday January 02 2019, @12:01PM
The thing is that <insert name or category of publication here> writes at its readers' level of comprehension and not necessarily in the industry's nomenclature.
Our designer (who also functions as our editor for all public facing content) reminds me of this all the time. As a programmer I naturally write content for other programmers to read rather than clients (or prospects). Our designer highlights things she feels are inaccessible to those outside of our industry so I can refine (or rewrite) them.
(Score: 2) by zocalo on Wednesday January 02 2019, @01:32PM
Finally, to clarify, I cited the DLR becase, unlike the Victoria Line (and many others since), the DLR trains are unmanned (other than an optional guard that wanders around the train), and it's a bit more complex than most unmanned airport systems where trains just shuttle back and forth on a single route or circle around a loop. All DLR train control, including movement across various interchanges and multiple branch lines, is handled using a combination of local sensors and central management, just like Rio Tinto's system. There's always an element of bespoke in this kind of project, but surely AUD $1.3b has to buy you more than smarter engines, trackside sensors, and a fancy control room?
UNIX? They're not even circumcised! Savages!
(Score: 5, Insightful) by Anonymous Coward on Wednesday January 02 2019, @10:33AM (2 children)
We're engineers here (well, at least *I* am, don't know about you). So let's not go the sheeple SJW route but be professional and do a rough calculation.
The 2018 average full time salary in West Australia was 90,000AUD. The average full time salary in the transport industry (for all of Australia) was also 90,000AUD. This sounds reasonable to me, so let's use that number. In western countries, normal workers typically cause total costs of about twice their income, so 180,000AUD per workeryear. So the breakeven point for the initial investment is at 7220 workeryears. Spread over 14 mines and 20 years of operation, the breakeven point for the initial investment is at 26 transportation workers per mine.
I do not think they have that many train drivers to make the automation worthwhile, my personal guess would be 3-5 train drivers per mine (presuming that loading/unloading, already highly automated, will remain as is). This calculation does also not include the operating costs of the new system. These are hard to judge from the information we have: less workers, but more skilled. Also, direct and indirect costs for software maintenance may be high during the first 5 years of operation.
So no, it's probably not just the personnel savings. For my above guess, these would amount to 250 million total for 20 years. Only 20% of the initial investment. OK, so we're missing something big here.
I then did a very short web search for "west australia mining train" and immediately found the culprit: averaging over 2015-2018, the West Australian mining company BHP had one ore train derailment per year, all of them in the boonies without fatalities. For the 2018 incident, the company stated a repair time of about a week while an unnamed "industry source" guesstimated an operational loss of 55 million AUD per day for that time. That's about 400 million AUD for the total incident, including the cost of replacing a "smaller" ore train and hiring costs for a new train driver. Reasons for this incident are unclear: "train started to move while driver was out doing checks", then continued driving for an hour and was purposely derailed to protect population (or the loading terminal ....). So we can safely assume it will be found (if not proven) to have been some kind of human error.
For the same 20 year amortization period as above, that amounts to 8 billion AUD in costs. This is probably a conservative number, considering that other derailments were bigger, and that some day a derailment may happen in a town. Compare that to the 1.3 billion for the new system, plus the new system's operation costs, minus 250 million for saved labour, and you have an immediate, juicy business case.
Yes, the new system will have troubles, and these will cost money. But unless the robot produces more derailments than the humans, you still have a winner. Multiply the human condition with 400 mega-AUD per realized fault, and there's your reason for a few jobless train drivers.
Of course, instituting _other_ measures might also have lowered the derailment rate. I do, however, assume that these other options have been studied and found lacking for some reason or other.
(Score: 0) by Anonymous Coward on Wednesday January 02 2019, @12:04PM
This is a great post.
(Score: 1) by anubi on Wednesday January 02 2019, @11:06PM
We haven't had elevator operators since I was a kid, and that was well over fifty years ago.
Now, I see a train as more like a horizontal elevator, confined on rails. It is extremely limited by its construction what it can do. It puzzles me why a human is involved in it anymore at all. This is a job for relays, motors, and comm lines to report malfunctions.
"Prove all things; hold fast that which is good." [KJV: I Thessalonians 5:21]
(Score: 3, Insightful) by c0lo on Wednesday January 02 2019, @10:44AM
Other than avoid hiring those expensive flyovers and outsource the tele-operation of those mines in India? I'm afraid I can't do that, Michael.
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
(Score: 1) by khallow on Thursday January 03 2019, @03:10AM (3 children)
But yea, spending over a billion dollars to remove some human engineers might indicate something about Australia's incentives for employing humans.
(Score: 2) by c0lo on Thursday January 03 2019, @06:46AM (2 children)
Yes, unusually high temperatures for many days in a row (most recent: end week of last year at over 45C, with 2-3 days in a row with temperatures hovering 49C).
You have to pay well into 6digit for the miners to fly-in-fly-out in Pilbara.
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
(Score: 1) by khallow on Thursday January 03 2019, @01:35PM (1 child)
There's always AC. Given the expenses, they can always add AC to the quarters of the engineers.
Sure, but they also do more than just sweat when they're out there. I imagine the pay is pretty good for people who carefully watch over 2+km trains loaded with refined ore. The problem I was referring to is that even with that, it's quite easy for governments to add substantially to the costs of employing people, such as taxes, "superannuation" [ato.gov.au] (retirement "savings"), working "entitlements" [ato.gov.au], fringe benefit taxes, and a host of regulations to show compliance with (which means HR overhead).
(Score: 0) by Anonymous Coward on Thursday January 03 2019, @11:40PM
Oh, what has come to this world if neither global corporations nor govts listen to khallow's wisdom.