Announcement

Collapse
No announcement yet.

Bridge Crane Detection Sensing

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts


  • Bridge Crane Detection Sensing

    I have a project where a bridge crane has the potential to collide with a machine below while it is in operation. I need to detect two different things.

    1. Some sort of collision sensor on the bridge to detect when the machine is in it's path. The bridge span is approximately 100' and the collision zone is probably 20' of the bridge.
    2. A way to detect the presence of the crane to prevent the machine from operating when it's in the way. The collision area spans over 150' of runway length. Basically, the crane can't be in this 150' "zone". If it is, the machine below is inhibited from operating.

    The duty cycle of the machine is low (about 3 cycles in a 24 hour period with each cycle being less than 5 minutes), but there are multiple machines under the same crane. In other words, the same crane has to avoid 2-3 of these 150' zones at different times of the day.

    #1 is probably the easier one to solve. I know they make different types of sensors to prevent cranes from colliding when there are multiple cranes on the same runway. I can probably adapt one of these to look for an object in it's path.

    #2 has me stumped. I can't think of anything short of a 150' long array of sensors along the runway where any sensor would inhibit the machine below.


  • #2
    1) You could mount a safety area detection sensor on the crane. I think Sick and Idec have those, maybe Banner. banner also has a radar that I think will go 100 ft

    2) I assume the crane is manual and doesn't know its position. If it does, use that to create zones with output contacts for each zone. If it does not, then use a laser rangefinder looking at a reflector on the crane. Then take that analog and put it into a controller that will output an inhibit contact DO for each potential conflict zone. Or use one of the sensors from (1) for each machine, looking for the crane to be in its zone.

    Comment



    • #3
      Originally posted by ControlsGuy View Post
      I assume the crane is manual and doesn't know its position. If it does, use that to create zones with output contacts for each zone. If it does not, then use a laser rangefinder looking at a reflector on the crane. Then take that analog and put it into a controller that will output an inhibit contact DO for each potential conflict zone. Or use one of the sensors from (1) for each machine, looking for the crane to be in its zone.
      It is manual (has no idea where it is on the runway). What range is reasonable for the laser range finder? The crane runway is over 1000' long.

      Comment



      • #4
        I found these guys by Googling, but have no direct experience with them http://www.lasertech.com/TruPulse-La...ngefinder.aspx It says they can go 2km with a reflector or 1km without, either of which covers you.

        I've used ones recently from Keyence and they worked well, but I don't think their max range goes that high.

        Back in the 80's we used a unit from an outfit called TeleControls Systems, iirc, and they were previously an HP product used for surveying, so I'm sure they'd have to be able to go that far, though we were using them on automated rigid cranes with max travel of about 500 ft.

        Oh, and I meant to mention in my original post that depending on duty cycle you could delineate the zones with flags and maintained limit switches (two per zone). The switches have a fork-shaped actuator and flip into one of two maintained positions when you pass the flag. That position puts the roller on the other side of the fork into position so that when you again hit the flag traveling in the other direction, it flips it back. So two of those, one to sense the max end of Zone X, and one to send the min end, and wire them in series, and you have a contact that says you're in the Zone. You could do it with a single flag on the crane and two switches per zone.

        You could also do a long linear unrolled cam box, with normal non-maintained limit switches (or proxes), but then the cam has to be the entire length of the zone, so might be more work than it's worth.

        Remtron has a lot of good crane controls (with safety built in) so you might look at them.

        Comment



        • #5
          I like your limit switch idea, but there are multiple cranes on each runway (sorry, forgot to add that in the OP). A variation of it might work though. What if prox sensors were at each end of the zone triggered by a target on the cranes. It would take two at each end to determine direction, but this could be used to keep track of how many cranes enter and exit the zone. This would also keep the controls local to the zone so we wouldn't have to run 100's of feet of control wire to a sensor at the end of the building. Thoughts?

          Comment



          • #6
            That's pretty slick! And I like the idea of keeping the wiring associated with the zone, both logically and wiring cost wise. My concern about proxes is that your system would be sensitive to an intermittent signal (fooling the controller about the order of loss, especially if the paired proxes are close and their signals overlap) and I wonder if the cranes' alignment is good enough to ensure no signal flicker. Maybe they are.

            The other thing I'm not crazy about is having to remember where the cranes are. What happens if the PLC's blow a fuse or something? I'd prefer something that tells you in real time if there's a crane in the shared space. What about a long range through beam photo looking diagonally across the crash zone? Mount it at an elevation where the crane is guaranteed to block the beam no matter where it is on the runway within the crash zone. Plus, since the photos have to see to permit machine movement, you pick up some fail-safety.

            Comment



            • #7
              Possibly a Local Positioning System, Like GPS for indoors...

              http://www.itrack-llc.com/download/lpts.pdf

              Comment



              • #8
                Originally posted by ControlsGuy View Post
                My concern about proxes is that your system would be sensitive to an intermittent signal (fooling the controller about the order of loss, especially if the paired proxes are close and their signals overlap) and I wonder if the cranes' alignment is good enough to ensure no signal flicker.
                I think I could manage the alignment by sensing in a plane where the distance doesn't change (like mounting the sensors vertically) and properly sizing the sensors and targets. I think you do bring up a valid concern.. maybe some sort of redundancy?

                Originally posted by ControlsGuy View Post
                The other thing I'm not crazy about is having to remember where the cranes are. What happens if the PLC's blow a fuse or something?
                That's a concern, but pretty low risk I think. It's going to need way to override/reset the crane counting register anyway. That said, having people that know how this works could be an issue if it screws up at 3:00 AM.

                Originally posted by ControlsGuy View Post
                I'd prefer something that tells you in real time if there's a crane in the shared space. What about a long range through beam photo looking diagonally across the crash zone? Mount it at an elevation where the crane is guaranteed to block the beam no matter where it is on the runway within the crash zone. Plus, since the photos have to see to permit machine movement, you pick up some fail-safety.
                This is an interesting idea. It's both real-time and local. The sensor and target (or emitter and receiver) would be at least 180' apart. What kind of sensor has that kind of range (and how would you aim it )? I'd also be a little concerned about vibration and dust (weakening the signal).


                Comment



                • #9
                  Originally posted by ljbeng View Post
                  Possibly a Local Positioning System, Like GPS for indoors...

                  http://www.itrack-llc.com/download/lpts.pdf

                  Interesting. I perused their site a little and found a couple of things that might work. I may reach out to them and see how viable (and expensive) this is.

                  Comment

                  Working...
                  X