Designing a climbing mechanism for 2013... a humbling experience

[Nemo]

Originally Posted by Chris is me

Second - if you throw 5 weeks into a 30 point hanger because "team X will do a 30 point hang AND shoot frisbees", and you don't use it, that's hundreds of man-hours you could have put into your shooter, your intake, or into drive practice in order to get your scoring rate up at or above their level! The jack of all trades really is the master of none for this game more than many others. And if 1625 matches you frisbee for frisbee and can 30 point hang: assemble a better alliance and the match is still in question. Three specialized robots will often beat one multi-function robot and two supporters.

This is a good post, and I have pondered it off and on during the past day.

One thing I'd like to add is that we try to evaluate our own abilities as we choose our strategy. We are aiming for something that does more than a minimum competitive concept, but not necessarily everything - we believe we are capable of that. We want to design a robot that can win the regional, rather than getting into elimination rounds as a middle alliance member and then facing long odds against the #1 alliance. If you're trying to win, I think you need to make your best educated guesses about what the competition is going to bring and design accordingly. With that in mind, I think it would be great to quickly finish a bot that shoots, human loads, and does a 10 point hang. But I don't think that is enough to actually win the competition, so I think we need at least one more ability on top of that.

I am still not certain that we are capable of a 30 point hanger, but our progress seems promising enough to continue pursuing it, because the payoff is pretty substantial. We aren't putting all of our eggs in that basket - we are developing a more frisbee oriented robot design as well. Adding a 10 point hanger as a substitute will not be very difficult if we decide to go that route. Splitting our resources isn't ideal, and deciding which route to take will be something we constantly evaluate in this segment of the season.

I'll say this much - this year's game is forcing some really tough choices. I completely agree that it doesn't make sense to try to do everything.

Edit: One other thing I was going to mention - Yeah, I do have a healthy amount of fear that spending too much time on a climber and then having to scrap it will result in a rough season. It's a risk vs reward thing that keeps me awake at night.

[alicen]

although I'm not working with a team this year (reffing at Peachtree) I have given some thought to a simple climbing device. It would allow for relatively simple removal as well. I'll have to make some sketches of it tomorrow and post it up here

I'll give a quick shot at explaining it though.

Picture a wheel, it has no hub, but is still driven. (This can be done) Now imagine it has "fingers" coming off of it, kind of like hooks that would hold the corner of the tower and push it against the wheel. These fingers would alternate which side they held the tower from (left, right, left, etc) and would connect and disconnect by a cam that would push them out (to allow it around the tower corner pole) and then spring force, or surgical tubing, etc. would snap it back, pinching it to the wheel. As the wheel rotates, it simply drives right up the tower, each of its little sloth fingers holding it in place.

Of course this idea isn't remotely fleshed out and would need something like skids to keep the robot from slipping and rotating around (unless that was all part of the plan!), but that is my attempt at relatively simple and reliable

[rinim324]

Originally Posted by MrForbes

Do you want an opinion on the legality of it? or whether or not such a mechanism could work?

Either way, we'd need a lot more information about the design.

Whether or not it would work. We actually revised it a little to have 3 stages and a slider peice on the end with hooks attached to it. The robot is designed to be a corner climber. The arm fully retracted fits inside the robot frame but fully extended can reach 38inches. The Arm is fixed on the frame at 60 degrees (The corner posts are 60 degrees relative to the floor not 68 like the sides). We extend the lift arm with a simple pulley mechanism but retract it with a winch that has the cable hooked to the outermost stage and pulls down at 60 dgrees parallel to the The Arm. We also have a wheel mechanism at the bottom of the robot that rides the vertical post as a third point of contact and it is designed to sort of "walk" over the corners of the horizontal bars. Once the arm is retracted fully two sets of hooks mounted to the frame will latch on to the horizontal bar while the arm extends to the next level.

I hope this is the info you would need.

[JamesCH95]

Originally Posted by Kevin Sevcik

So just to keep everyone in the thread updated, Team Update 1-15-2012 has suddenly made climbing a rather lot easier. The 54" cylinder is now robot relative while you're touching the pyramid. Which obviously makes many mechanisms and methods a lot more viable.

My condolences to those of you who have already locked in a more restricted, less robust design based on yesterday's rules.b

In your opinion the rules forced a less robust design. I disagree. We had a nice robust design that even the most conservative members of our team were happy with. Now it's illegal.

[EricPalmatier]

Originally Posted by Nemo

Edit: One other thing I was going to mention - Yeah, I do have a healthy amount of fear that spending too much time on a climber and then having to scrap it will result in a rough season. It's a risk vs reward thing that keeps me awake at night.

I totally agree, except add to that the fear of watching your robot fall from L3 of the pyramid. That would be a game ender for most robots, and that is what has been keeping ME awake at night. I totally think that it's an achievable task, albeit extremely challenging. However, the GDC wouldn't ask the FIRST community to tackle a problem that they themselves Have not already tackled (to a certain extent), and did not think we could handle.

As to the fear of spending too much time working on a L3 system. Again, I totally agree that it is a HIGH risk and HIGH reward scenario. To devote a serious amount of time and effort working through the problem and fabricating prototypes, just to figure out that for whatever reason ( mechanism doesn't work, weight budget allotment, team change in strategy, system complexity and/or cost) would be a crushing blow to not only the individuals directly involved but to the teams ability to tackle the other facets of the game.

With that said, I know my team has been disproportionately allocating mentor/student resources towards frisbee shooting/collecting and general manipulation as opposed to the climb. It's a safer strategy, and if you run the numbers, an effective shooter can definitely negate a good climber. Now, me personally, believe that climbing is the more interesting challenge and as a result I'm the lone mentor on my team (crazy enough, I guess ) to attempt to tackle this problem; and even if I can do it, it will be a tough sell. Again, due to the risk.

I know other teams are dealing with the negative responses to the ideas of climbing, has anyone come up with a valid way to sell a potentially robot destroying strategy, or is the safety inherent in the design? In other words, the safety features of the design / mechanism sells the team on it's feasibility.? Just wondering if anyone else is getting resistance and how you are dealing with it.

Finally, my first climbing prototype, which is no longer the way I want to go about it, had quite the mechanical failure. I thought I would share this to illustrate just what kinds of forces you are dealing with when trying to use an arm to 'curl' the weight of the robot. The shaft was a .5" 4140 steel shaft, acting as the output of a 5-stage AndyMark GEM planetary gearbox giving us a reduction of 666:1. Perhaps I should of known it was destined for failure based off of the reduction. (note: it curled 178 pounds before failing )

EDIT: And yes, this challenge has completely humbled me!

[Jake177]

Someone just posted this in the thread specific to Team Update 3.
http://www.chiefdelphi.com/forums/sh...&postcount=110

It appears that anything which was legal under the original rule/Q&A responses is still legal.

[Djur]

Originally Posted by EricPalmatier

Finally, my first climbing prototype, which is no longer the way I want to go about it, had quite the mechanical failure. I thought I would share this to illustrate just what kinds of forces you are dealing with when trying to use an arm to 'curl' the weight of the robot. The shaft was a .5" 4140 steel shaft, acting as the output of a 5-stage AndyMark GEM planetary gearbox giving us a reduction of 666:1. Perhaps I should of known it was destined for failure based off of the reduction. (note: it curled 178 pounds before failing)

What was the output attached to?

[Mr. P]

I would like to see team 842s climbing storyboard. Is there a link?

[2348humanplayer]

We have made one thet could work. Not sayinig how, but I'm just gonna say one way valves.

[MattDAngeli]

It's not going to be easy at all. We are still brainstorming ideas to climb, and it's still not looking good, considering we want a robot that does all.

[s_forbes]

Originally Posted by Mr. P

I would like to see team 842s climbing storyboard. Is there a link?

You can see a document tracking our build season progress here. There is a climbing sequence about halfway down that shows the concept we are planning to use.

[wireties]

Perhaps a heavy lift quad-copter that reaches out and taps each horizontal pipe, dumps 4 blue/red discs and hovers in level 3? ;o)

[Cal578]

Originally Posted by rinim324

Can i get an expert opinion on a design that uses a three stage arm which extends via a pulley system but retracts using a winch powered by a toughbox? The Stages are 27in long with 8in of overlap.

Originally Posted by rinim324

Whether or not it would work. We actually revised it a little to have 3 stages and a slider peice on the end with hooks attached to it. The robot is designed to be a corner climber. The arm fully retracted fits inside the robot frame but fully extended can reach 38inches. The Arm is fixed on the frame at 60 degrees (The corner posts are 60 degrees relative to the floor not 68 like the sides). We extend the lift arm with a simple pulley mechanism but retract it with a winch that has the cable hooked to the outermost stage and pulls down at 60 dgrees parallel to the The Arm. We also have a wheel mechanism at the bottom of the robot that rides the vertical post as a third point of contact and it is designed to sort of "walk" over the corners of the horizontal bars. Once the arm is retracted fully two sets of hooks mounted to the frame will latch on to the horizontal bar while the arm extends to the next level.

I hope this is the info you would need.

This is an interesting concept, and has similarities to our plan.

By "3 stages" do you mean 3 parts that move, in addition to the fixed end? That is complexity that can be difficult to put into operation. Depending on how you use the pulleys to move the last stage, that could add up to a lot of torque on the first stage. Without seeing a drawing, that would be hard for us to evaluate. But it sounds good, and I hope it works for you.

[Cal578]

Originally Posted by wireties

Perhaps a heavy lift quad-copter that reaches out and taps each horizontal pipe, dumps 4 blue/red discs and hovers in level 3? ;o)

Other than safety issues, if you could make that work within all the robot rules (like energy source), I think that would be perfectly legal, and a valid climb. When you're done, don't forget to post the video.

[Mr. P]

Thanks, you guys are doing a great job!