Student Robotics - Carrying cardboard boxes

Thread Starter

tom66

Joined May 9, 2009
2,595
I have joined the robotics club at my college.

We are entering a robot into the Student Robotics competition, hosted at Southampton University. We are part of team QMC and we have a little under 5 months to complete the project.

The challenge is to make autonomous robots which pick up cardboard boxes.

So I am trying to think of a way to pick up the boxes. I was thinking of compressed air. So how to provide the necessary suction? A bicycle tyre pump came to mind as one option - but how to actuate it? Solenoid? Central locking actuator, like this? A motor? An electric pump, like this? But weight and power consumption is important...

Also, the suction needs to be maintained for at least 30 seconds while the box is carried, even if the robot moves around or bumps into things. Then, when ready, there needs to be a way to release the box.

If anyone has any ideas, I would appreciate input. Thanks.
 

Wendy

Joined Mar 24, 2008
23,408
If you need a vacuum you could use something like a small hand held vacuum, complete with battery pack that is included. All you'd need to do is add a switch across the build in switch.
 

Thread Starter

tom66

Joined May 9, 2009
2,595
If you need a vacuum you could use something like a small hand held vacuum, complete with battery pack that is included. All you'd need to do is add a switch across the build in switch.
A vacuum is a good idea - but I wonder if it provides enough suction?

One thing I forgot to mention - the suction should be reversible. i.e. the electric tyre pump should be able to suck as well as the default blow. I expect it's just a simple brushed DC motor inside in which case reversing the terminals with a changeover relay, or a H-bridge, will make it act in reverse.

Is there any way for me to bottle up this pressure, then suddenly release it? It would be great if the boxes could be sent flying into the home zone, that way the robot would not even need to move that much.
 

Wendy

Joined Mar 24, 2008
23,408
You don't need to reverse suction, just turn it off. Leaks will take care of the rest.

As for the strength, that is a function of cross surface area. You could likely warp the cardboard if the vacuum orifice was too large. I am assuming you have a flat surface to work with.
 

ErnieM

Joined Apr 24, 2011
8,377
Suction seems a little risky as towards the end of the competition the boxes may get squashed and thus not have a nice planar surface to suck on.

How about just tongs that compress like a thumb and for finger to grab it? That's all electric control for little power.

These things tend to get won when you think outside the box: What if you harpoon the box and release it in your goal?
 

Thread Starter

tom66

Joined May 9, 2009
2,595
Good point, what do the rules say about damaging the boxes?
The boxes must not be damaged! :(

Last year (I wasn't involved) team QMC went for a grabber. (The goal was to pick up full cans of beans.) That worked, until another robot accidentally collided into it, and damaged the mechanism.
 

shortbus

Joined Sep 30, 2009
10,045
Cardboard is fairly porous, so suction may not be the best option. A vacuum pump or for that matter an air compressor doesn't usually change it's function by reversing its motor. Its all in the way the valves in the pump are biased.

Are the boxes on the floor or on a table? A way that may work is to have a thin plate, say around 1/8" thick to slide under the box and then lift up. Kind of a robotic fork-lift.
 

Thread Starter

tom66

Joined May 9, 2009
2,595
Cardboard is fairly porous, so suction may not be the best option. A vacuum pump or for that matter an air compressor doesn't usually change it's function by reversing its motor. Its all in the way the valves in the pump are biased.

Are the boxes on the floor or on a table? A way that may work is to have a thin plate, say around 1/8" thick to slide under the box and then lift up. Kind of a robotic fork-lift.
They are on a floor; the floor is carpeted.

We have been given some sample boxes, so I'll have a look at them.

So are you saying that tyre inflater wouldn't suck air if it were reversed? Does this mean I need another way to generate suction?

I was also thinking of how to make the boxes fly out. I'd need a way to send the boxes flying at 15 m/s at a 10 degree inclination (There are lots of other combinations that work, and the velocity/angle also depends on how far we need to send it flying, but assume at least 5m would be best.) This would cause it to rise to 0.39m (including a 0.05m starting height) and travel 5.31m, which would get it within the zone. The boxes must not touch the ceiling at 500mm high, so the angle is set low.

To achieve the suction, I could just use a rubber suction cup, but I don't know if it provides enough suction. Then we could bottle up the pressure from the tyre inflater into a cola bottle. A valve or actuator can then be used to instantaneously release this pressure, and send the cardboard box flying. The right pressure would need to be used (and calculated in software); measuring pressure in the bottle could be achieved by measuring the inflation time (experimentally.)

The robots are autonomous so a bit of game logic will need to be devised. The robot has a web cam and a BeagleBoard on it to capture high resolution video and process it to examine the field for markers on the objects. Each object has a sort of bar code which helps the robot see its world. The vision could be supplemented with micro-switches or sonar.
 

shortbus

Joined Sep 30, 2009
10,045
What is the weight of said boxes? You say they must be thrown? Do they need to hit a target area?

With the 'fork-lift you could also incorporate a 'flipper'. The lift would go up, with the box on it, as it went up it would tension a extension spring that would then power the flipper.

Can you post the complete rules for your contest?

Most small air compressors just have disc valves in them to control the direction of air flow. Doesn't matter which way the motor turns, the air comes in on a piston down stoke and out on the up stroke. Guess you could put a tee and two solenoid valves to choose whether air came in from a suction cup or normal room air, never tried it .
 
Last edited:

Thread Starter

tom66

Joined May 9, 2009
2,595
What is the weight of said boxes? You say they must be thrown? Do they need to hit a target area?

With the 'fork-lift you could also incorporate a 'flipper'. The lift would go up, with the box on it, as it went up it would tension a extension spring that would then power the flipper.

Can you post the complete rules for you contest?
The boxes are cardboard, so I'll assume the weight is "somewhat" insignificant, probably 10-20 grams each.

Yes the landing location is important. The robot might be in a specific position in the room and needs to fire the block back to its zone. The zone is a 1m x ~8m area and has a wall behind it, so it doesn't matter *too* much if we fire it too far; the wall should cause it to bounce (provided, of course, the box doesn't bounce back out of the area.)

The boxes do not need to be thrown, that is just one potential strategy. They could be picked up and deposited if required, but I think it would require more time and thus get a lower score. The goal is to get the most boxes into the zones. Also, there will be three other robots competing in the arena, so the faster we are the higher the score.

A flipper is another option, but would it provide enough velocity in a short travel to send the box flying to the right area?

Rules are here: https://www.studentrobotics.org/docs/rules/
 

Thread Starter

tom66

Joined May 9, 2009
2,595
Most small air compressors just have disc valves in them to control the direction of air flow. Doesn't matter which way the motor turns, the air comes in on a piston down stoke and out on the up stroke. Guess you could put a tee and two solenoid valves to choose whether air came in from a suction cup or normal room air, never tried it .
So, a tyre inflater is a good source of pumped air, but not so good for suction. I'm confused about what you mean by piston? I thought these were just electric motors and impellers inside a casing, hence the 12V operation. Maybe I could remove the valve?

Also, I tested cardboard with a suction cup - no luck. Printed cardboard is okay, but the bare stuff doesn't stick.
 

Georacer

Joined Nov 25, 2009
5,182
11cmx11cm seems to be a rather small size for a box. Why not use a common grabbing technique? Something like two or two pairs of fingers, for example.
 

shortbus

Joined Sep 30, 2009
10,045
Here in the states they are just a small compressor, A crank hooked to the motor shaft, a cylinder, a piston and a cylinder head with the valves in it. Without valves it would just oscillate the air back and forth. The valves allow the compression of the air.

A vane type pump is often used for vacuum but not good for compression. Too much air slippage past the vanes when pressure starts to build. Unless some type of axial vanes are used.

I'll look at your rules and maybe can come up with other ideas/suggestions, if thats OK? When still working, I was involved with a robot contest similar to this making mechanical bits for the students.
 

Thread Starter

tom66

Joined May 9, 2009
2,595
Here in the states they are just a small compressor, A crank hooked to the motor shaft, a cylinder, a piston and a cylinder head with the valves in it. Without valves it would just oscillate the air back and forth. The valves allow the compression of the air.

A vane type pump is often used for vacuum but not good for compression. Too much air slippage past the vanes when pressure starts to build. Unless some type of axial vanes are used.

I'll look at your rules and maybe can come up with other ideas/suggestions, if thats OK? When still working, I was involved with a robot contest similar to this making mechanical bits for the students.
Thanks for your ideas.

I am meeting up Friday with the team again. I will submit some of my suggestions, as well as some of yours.

How would you suggest I build up some pressure to send the blocks flying?

Let's say the robot picks them up without suction (some other way, like a grabber) but then we want to catapult them... how best to do that? I'm convinced this would get us the highest score - as most of the other robots will be picking up the blocks and depositing them in the zones - so throwing or catapulting the blocks will give us a significant advantage.
 

Georacer

Joined Nov 25, 2009
5,182
You have less than 46cm of vertical spacing to propel your box over and into the basket. Let's keep in mind that carton boxes tend to bounce on surfaces they hit. If you manage to calculate the throwing force to make it just get over the basket from any shooting location, only then I can see how this could work. If you think you can program and take into account all these parameters, then go for it, you are the man.

Throwing objects is tricky, to say the least. I 'd say do some testing and if you see it's taking too much developing time, drop it, and go for reliability.

About the throwing method, since you don't have too much space overhead, you could build a crossbow-style slingshot, maybe with some sort of bungee cord. The trick would be to place the cubic box exactly parallel in the sling rail.
 

Thread Starter

tom66

Joined May 9, 2009
2,595
You have less than 46cm of vertical spacing to propel your box over and into the basket. Let's keep in mind that carton boxes tend to bounce on surfaces they hit. If you manage to calculate the throwing force to make it just get over the basket from any shooting location, only then I can see how this could work. If you think you can program and take into account all these parameters, then go for it, you are the man.

Throwing objects is tricky, to say the least. I 'd say do some testing and if you see it's taking too much developing time, drop it, and go for reliability.

About the throwing method, since you don't have too much space overhead, you could build a crossbow-style slingshot, maybe with some sort of bungee cord. The trick would be to place the cubic box exactly parallel in the sling rail.
We don't *have* to get the boxes in the buckets, but that gives us bonus points. Getting them in the zone gives 2 points, in the bucket 5 points.

I was thinking that programming the robot to gather many boxes into the zone, and with 1 minute left on the clock, get as many into the ship (bucket) as possible. Or perhaps when N boxes have been collected, go and put them in the ship. If it takes 15 seconds to pick up a box and throw it in the zone, but 30 seconds to get it in a ship then the extra 3 points aren't really worth it - but we'll need to actually test this to be sure.

Crossbow is a good idea.

I think as the environment has virtually zero wind (indoors) simple trajectory equations could be used.

I wrote a simple spreadsheet in OpenOffice to calculate trajectories... attached (rename it to an *.xsv.) Plug a few values into the equations and we can see that it's not too difficult to get a long distance flight, but we do need a lot of velocity (and hence, a lot of pressure.) 15m/s = ~33.6mph.

The problem is reaching that in just 5m. Acceleration needs to be calculated.

v^2 = u^2 + 2as: u^2 = 0 (initial velocity), v^2 = 100, a = ?, s = 5m.

Solving for a gives us 22.5ms^-2. Force = ma; if we say a mass of 30g for a box (a guess) the force is 0.67N.

If we apply pressure = force / area, and an area of a 2cm diameter circle (an assumption for our "sucker"), area = 0.063m^2. Pressure = 10.6Nm^-2 = 10.6Pa = 0.0015 psi.

That seems too low, have I gone wrong?
 

Attachments

Last edited:

thatoneguy

Joined Feb 19, 2009
6,359
Are pyrotechnics allowed?

I'm thinking, for box ejection, one of those blanks used in nail guns. They are very consistent, so once you have the distance figure out, bang, and it's in the bucket. You'd need to use a pusher plate to keep from damaging the box is all.

So the boxes are only a 4" cube? How thick are the walls? Any seams on the side that will be 'up'?
 
Top