The World's Heaviest Robot 142
Roland Piquepaille writes "This distinction goes to a future autonomous version of the 700-tons Caterpillar mining truck. In this article, Discovery News reports that Caterpillar engineers and computer scientists from Carnegie Mellon University have teamed up to develop this autonomous truck. Japan-based Komatsu has already delivered autonomous mining trucks to its customers, but these are smaller than the Caterpillar ones. Both companies are transforming their trucks into 'robots' for three reasons. Improvements in safety, efficiency and productivity will reduce costs and increase availability."
Heaviest at a mere 700T? (Score:5, Interesting)
I would say the distinction would go to Australia's driverless ore trains when commissioned in 2013.
http://en.wikipedia.org/wiki/Image:Port_Hedland,_Western_Australia.jpg [wikipedia.org]
(Yeah I know the pic is of a drivered BHP rather than Rio train... but it does give you an idea of the scale).
The argument for driverless is that because they take the best part of a day to perform an emergency stop, the family car stalled on the crossing is going to be pulverised regardless of the skill of a human operator. The largest fragment left over from these collisions is usually a few inches in size. Fortunately they don't happen that often.
Why don't rockets count ? (Score:3, Interesting)
An unmanned spacecraft when launched is "fully autonomous" and there are a number that are bigger than 700 metric tons.
Titan IVB - 943 tons
Delta IV - up to 733 tons
Saturn V (there were 3 unmanned launches) - 3000 tons
Re:Heaviest at a mere 700T? (Score:2, Interesting)
Typical Rio Tinto Pilbara train
Loaded weight 30 000 tonnes
Length 2.4 km
Top speed 75 km/h
http://www.railwaygazette.com/news_view/article/2008/10//rio_tinto_to_go_driverless.html
Mining in the Pilbara (Score:5, Interesting)
I live in Western Australia, where Rio Tinto are developing their autonomous mining equipment. My old man is employed by Rio as a radio techie and as such he has a small part to play in the development of their autonomous trucks. I'm posting anonymously as I don't want any of these remarks getting him in trouble.
The basic model relies on a single central system coordinating all the trucks as they move about. There have been a lot of kinks to do with radio lines of sight. Mine sites where the principal ore body is ferrous tend not to be very friendly to radio transmission. The actual driving of the trucks is not hugely complicated as the routes are identical for every run and GPS systems make it easy to align a truck with the road. It's not like navigating a desert track or urban landscape.
Rio are also working on automating their trains. Alongside BHP, Rio has one of the largest fully private rail networks in the world. The two mining giants run the largest freight trains anywhere -- several kilometres long with tens of thousands of tons of iron ore. That may change as the two mining giants are being forced to open their networks to junior mining companies -- it looks easy on paper but railway systems are extremely complicated and finely balanced systems.
The scale of the mining efforts in the Pilbara boggle the mind. They are utterly vast. Whole mountains made of rust are being excavated, crushed and then shipped to Japan and China. The pace is frenetic. Rio and BHP can't find enough workers and that has caused costs to explode, which is part of why they are turning to automation. But really it's about throughput: an autonomous truck doesn't need lunch breaks, delays for shift change or the like. Every minute counts.
Rio have been more forward-looking than BHP in this department, so it may not last as BHP are trying to carry out a hostile takeover at the moment.
Re:Mining in the Pilbara (Score:3, Interesting)
With a mine 2 kilometers deep and 7 KMs wide, you may well have LOS problems with LEO satellites! In fact, even GPS is a problem and most of these system don't use GPS at all, rather relying on a series of land-based transmitters on the edge of the excavation.
Re:Mining in the Pilbara (Score:3, Interesting)
To put some numbers on this a small ship loading plant hits an easy 300 megatonne per annum. The cost of downtime on a line is around $500 a second.
Re:Hit or not to hit, that is the question... (Score:3, Interesting)
That's about how it goes [putfile.com]
1 major problem: Robots are too accurate (Score:3, Interesting)
The biggest problem with the driverless Komatsu's on the mine sites are the ruts they cause on the dirt mine 'roads'. When you have 18 tippers weighing 500+ ton rolling on the exact same pair of tyre treads once a minute 24x7, the ruts get gouged pretty deep, pretty quick. A human driver will do his or her best to avoid ruts as he or she drives around every small (and large!) rock pile on the road as it makes for a smoother ride, especially when all you want to do is get it from the bottom of the pit to the top as many times as directed for your 8 hour shift.