Hi Raj, do you have any easier questions?? ;-)
Yes, GPS devices can supply 3D information but it will cost a packet. I
know of one instance of a Garmin GPS-35 (
http://www.garmin.com/products/gps35/ ) being used to supply 3D data to a
PLC system over a comms link, but have no specific references.
If I keep posting "easy" questions, then it would not be fun at all. I do
value the insights and thoughts provided by all of you over the past few
years (since I have been lurking on this newsgroup).
For your information (and other Aussie users), I have moved across from
Gladstone, Queensland (north east coast) to Perth, WA (south west coast).
The reason for my absence (a fortnight!!!!) is that all the furniture
(earthly possessions??) was in transit until now.
I worked on a project that had a "CAS" collision avoidance system that
controlled the speeds at which the various tools could move across the drill
floor on a semi-submersible oil rig.
As the operator moved a piece of equipment closer to another object, the cas
program would reduce the maximum speed available to the tool. Tools could
collide, they were just limited as to their speed at impact.
That said, the tool plcs used encoders (including angular encoders for the
booms) to determine their 3d location. I would be concerned how fast the end
of a boom could move under rotation and whether or not a refresh rate of 1
hz would be sufficient for that speed.
Also, the vertical accuracy of gps systems is not as good as the horizontal
accuracy. For the Garmin gps 35, with dgps, the horizontal accuracy is ~ 5
m. The question then is, will the error on two or more gps units on two or
more machines be the same and cancel each other out? or not? I think you
would want to vigorously test that on both vertical and horizontal
Just my two cents, Tom
Why not just use a local radio network. Having sensors embedded throughout
the area and transmitters on the stackers could let each know where the
other is (relatively), a software map of the area and some exchange of trip
data could let them negotiate optimal pathways to prevent them from running
into one another.
If you were only after collision avoidance, there are much better (read
simpler and cheaper) devices for doing this in an industrial environment -
eg. laser scanners from Sick www.sick.de Ultrasonic and radar scanners
exist too. You simply mount one on each boom and it will tell you the
moment anything comes within cooee.
We've used these gadgets on high-speed process cranes, and once installed,
collision avoidance was the least of our troubles..
To implement collision avoidance is more involved than implementing a zone
management system that absolutely prevents two machines from being in close
proximity. It has the benefit of allowing the machines to interact at a
slower speed. On the project I was working on, the scale of the machines,
and the large freedom of movement ( longitudinal, rotational, both vertical
and horizontal boom movement ) would have required a huge number of prox.
sensors. The software, also modeled the pipe the cranes were carrying which
could range in diameter from 6" to 4' with a length > 100'.
For a stacker system, zone management would probably suffice.
As Cameron understated, stackers don't move that fast. In fact,
stacker/reclaimers are very large machines that travel very, very slowly.
I don't expect they need an accuracy better that one meter,
probably less. They cost many millions and have proportional budgets
available. If they can eliminate a full time operator that would be
$500,000 per year. (One full time operator = four people on staff.)
It sounds like Raj is talking about a real-time mapping system. Accurate
computerized mapping systems exist. Our company has a subsidiary that deals
in such. I doubt anyone ever thought of doing this in real time to track
the changing landscape of a reclaimer pile.
news:J2Kdb.40551> As Cameron understated, stackers don't move that fast. In
As I had stated earlier, the system that I am familiar with, would allow
collisions, just at a reduced speed. In fact, one of the customer's client's
safety teams wanted a written guarentee that if a collission did occur,
nothing would break off any of the machines, fall fifty meters and kill or
maim a worker. As no one would sign such a document, the system was modified
to actually go to a max speed of 0% at the end of the prox. v. speed curve,
giving total collision protection.
My point is, if you are selling someone, an "anti-collision system"(Raj's
original query), it better work exactly as advertised, or you better have
great liability insurance. As such, I think you would want to test it
thoroughly, and document its shortcomings, if any.
Just because a machine is big and slow, doesn't mean you want it getting too
intimate with the neighbours. ;-)
Further to the below, I found an industrial GPS card for Allen-Bradley PLCs
(from Hiprom) that claim to be suitable for stacker-reclaimer systems. They
provide lat. long. and altitude or Cartesian coordinates as well as velocity
vectors to the plc backplane. Kinda neat, but they quote a non dgps Position
accuracy of better than 25 metres. A WAAS version would be really good for
North American applications.
Hiprom GPS card
wrote in message
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.