New to al this - large scale transporter

[Combat Wombat]

Hi,
am very new to all of this and not very electronically minded but have access to electronic engineers at work for explanations but can construct almost anything if I am told what is needed. What I am trying to work out is the construction of a system to move a platform type arrangement from point A to B, sense its arrival, then return to point A, and so on until either a time limit is reached or it runs out of power/switched off.
The platform will be carrying a large weight, say a maximum of approx 1000kilos (yes, 1000) and I envisage it being made out of tube/square steel and riding on car type tyres. Its for a display and we want to avoid rails, cables/pulleys, whinches etc and not require someone to sit there and watch it all day or control it by remote control. We also would like it to have the ability to stop if it senses something in front of it that shouldn't be there, hence the possible use of a commercially available sensor unit/suite to cut the drive power.
What I am curious about, could I also use a small commercially available robot such as the little surveyor robot sold by these folk
**broken link removed**
to drive larger servos to control larger heavier drive motors?
I know its a weird request - Oh, its not for DARPA/similar use, I am, in the wrong country for that and also I only want it to go in a straight line.......so far. Curves or bends are not in the plan until I can prove that it can go straight.
Any suggestions/advice?
Cheers

 

[user_88]

For the load that you have specified ... using pneumatic tires .... your wheel torque requirement will be several hundred foot-pounds.
One possibility is a hydraulic pump-motor system. There are fluidic gear pumps that can be powered by 12 VDC, as well as hydraulic gear motors, whose direction can be controlled with an electric valve. The reverse motion, emergency stop, and limit travel conditions could be designed into the system with some simple logic gates ... You might not require a micro-controller.
One important parameter would be the time duration for a given operating period, and the amount of electrical energy ... batteries ... that would be required onboard the vehicle.

Are there any other design requirements that would specifically require something like the robot you have mentioned?

 

[Combat Wombat]

I am guessing by design parameters you mean any other conditions/requirements?
If so, then no, I am guessing at the weight as I have not had the model made up (its huge) but am fairly confident that its weight as well as two smallish motors (based on the weight of car gear boxes with which I am familiar) and some (4) batteries woul dbe max 1000kgs.
If you mean have I seen anything like it? then, no as well. I spoke to an airfield electrician the other day and he mentioned the possibility of PLCs, that may be the way ahead if I can't adapt something else. Otherwise I might head into town to the Uni and speak to some 4th Year Eleetrical Engineering students who are doing robotics that I know via a fiend.

 

[dknguyen]

So you basically need a giant wheeled vehicle to move back and forth along a set path without any modifying the environment beforehand to serve as a navigational aids for long periods of time? You're going to need to at least need GPS. I think GPS accuracy should be good enough since it sounds big enough that the GPS error might not make too big a difference, but I don't know how accurate this path has to be. ANd if it's moving fast, you might also need an IMU to augment the GPS. Much like a UAV autopilot.

There's also the issue of safety for such a large heavy autonomous vehicle...

I do not recommend the Surveyor robot per se...this might serve you better...
The Surveyor brainboard:
Surveyor Corporation - Open Source Robotics
Autopilot Addon (for GPS and IMU...not too accurate but probably accurate enough for you)
**broken link removed**

You could probably just unplug the camera...I don't see you using that for anything. In fact, you might be able to just use the autopilot addon. IT does contain a dsPIC30 and can work alone. Not nearly as powerful as the blackfin, but you aren't exactly making a plane either so you might not need the extra processing power of the BlackFin. I'm not going to say that this is beyond all PLCs (beause I've never used oen so I don't know how powerful they are)...but you don't exactly need a PLC when you have a BlackFin or dsPIC already there.

I'm curious...it sounds like you plan to use electric motors? Where do you plan on getting the power from for the electric motors moving 1000kg back and forth all day long?

DISCLAIMER: I may be vastly over-estimating the distance that this thing has to travel which overcomplicates things.

 

[Combat Wombat]

IMU? sorry and very new to all this.
I thought we may need a GPS setup with it, but we were thinking of putting RFID tags at the perimiter to make sure it stays within a fixed area of ground - I agree about the potential for large scale carnage if it got out.
We could put other RFID tags on the ground at points to ensure it knew where it was in relation to a predeteremined path? Would that sound of use?

 

[bassnut]

This sounds like a standard industrial application engineering project, there are many off the shelf drive/motor solutions. The drive is not the hard part, liability and workplace safety requirements are. I wouldnt rely on students for help ;-). PLCs have many built in features to handle these type of applications which would be far more expensive to engineer from scratch with a PIC. And PLC power? hehe, only limited by dollars.

Your mucking around with serious gear here 1000kg self drive is no walk in the park, one hick up and ...................

 

[dknguyen]

IMU- inertial measurement unit. YOu know, accelerometers, gyros. WHat you do when you close your eyes while riding in a car (hopefully not driving the car!) and "feel" how far you have gone and how much you have turned, etc. to figure out where you are.

Really extravagant, yes...but it seems this little simple task of moving a vehicle back and forth autonmously without any landmarks or rails is on the low end of automated vehicles.

RFID tags are a good idea I think if you are able to get sufficient positional accuracy from them. Much cheaper to do and not so much work in preparing the environment.