Best Lift Design for Logomotion?

[nuggetsyl]

I will answer the question April 30, 2011

[Jack Jones]

The best I've seen so far is 469's (no surprise there). It weighs next to nothing and goes from the floor to the topmost peg in 0.6 seconds. It will be awesome as soon as they figure a way to control it.

[Alpha Beta]

What makes a lift system work well?

1. Robustness Some of the arms I have seen have a lot of lateral sway in them as the robot turns. Some elevators have weak attachments to the drive base. Both can be well built. If a robot is doing ground pick-up in traffic it will at some point have a manipulator outside of its frame where it can be hit. I would think the longer arms would sustain greater damage. (They also might be easier to repair and get back on the field though.)

2. Speed Arms and elevators seem equally fast to me. Arms may handle sudden changes in direction more poorly, especially if the gripper has significant mass at the end of the long swing arc.

3. Precision If done manually it is eaier for a driver to adjust the height of an elevator to find the peg than an arm. The movement arc is one dimensional instead of two. With driver practice, or software these issues can be mitigated.

4. Center of Gravity The weight stays more inside of the drive wheel in an elevator than it does for an arm. The s-forbes style arm is a nice way to avoid that problem.

I think elevators are better suited for this game, but not by much. Well built arms will beat poorly built elevators every time. The biggest difference will probably be in drive bases. The ability to get through defensive pressure and put the arm/elevator in a position to do its thing will be more important than the lifting system in general.

Just one man's opinion.

[MCahoon]

Originally Posted by Jack Jones

The best I've seen so far is 469's (no surprise there). It weighs next to nothing and goes from the floor to the topmost peg in 0.6 seconds. It will be awesome as soon as they figure a way to control it.

I haven't seen a reveal from 469. Do you have a link?

[JesseK]

Using what's been seen on CD and Youtube...

I think the 2-stage lifts* are working well this year. While slightly on the heavy side, they've been seen working with either bearings, custom linear slides (us), or 80/20 slides. They've varied in width relative to the robot, and in speed based upon how they're driven. They also have the most robustness in how they're driven -- pneumatic cyllinders, winches, and threaded rods have all been posted on CD. IMO, lifts are also easier to hold in position than an arm (unless ballast + window motors are used on the arm).

So for "best" from a performance perspective will be hard to tell (and will not be a dominating factor so long as it works well), even after the competition season. Yet "best" from a hollistic perspective is about dead even, given the variety of designs seen on CD.

*(2 stages that move + 1 base stage, it isn't a box tube within a box tube telescope setup, and the rotary joint for an end effector is on the upper stage)

[TEE]

I personally like 148's arm. It only uses one banebot motor, and it goes from all the way down to all the way up in a respectable time. They used a mechanical linkage rather than a chain

[Wayne TenBrink]

This thread is a great lesson about the design process - if you can figure out how to apply it. The original question was "what is best"? It turned into "what are we doing". The fact is, we don't what is best - yet. In time we will.

Early in the build season each team goes through its own process to estimate "what is best" and the result is "what we are doing". This happens every year. Only the mechanism in question changes.

We all apply our experience, think about how the game will play out, and judge our ability to actually produce the mechanisms we dream up with the available resources and time. How well did you document your thought process? Did you keep a record of your brainstorming ideas, and why/how you downselected to your final design approach? How well were you able to execute your design?

In you preparations for next season, go back through you design process for LogoMotion and see where you got it right and wrong. What timeless principles can you apply to repeat your successes and avoid your mistakes?

As for our team, we went with an elevator design. It is mechanically simple, quick, sturdy (doesn't wobble), easy for the drivers to line up under pressure, keeps critical hardware inside the bumper perimeter as much as possible, and is simpler from a controls (sensors and programming) standpoint. Each of these principles applies to just about every mechanism and every game. Did we get it right? Time will tell.

[Chris Hibner]

Originally Posted by Jack Jones

The best I've seen so far is 469's (no surprise there). It weighs next to nothing and goes from the floor to the topmost peg in 0.6 seconds. It will be awesome as soon as they figure a way to control it.

Interesting - I can't wait to see it.

Our elevator could get from bottom to top in less than a second, and we could do it with incredible precision. The problem was the forces generated by all that acceleration get pretty high. We now have it slowed down to about 50% of it's top speed, which is more than fast enough, but the power is there in reserve if it needs it.

I can't wait to see all of the robots this weekend!

[JVN]

Originally Posted by Chris Hibner

Our elevator could get from bottom to top in less than a second, and we could do it with incredible precision. The problem was the forces generated by all that acceleration get pretty high. We now have it slowed down to about 50% of it's top speed, which is more than fast enough, but the power is there in reserve if it needs it.

Bingo. We slow ours WAY down to make it useable.

These tubes are pretty light. It isn't hard to get them from floor to ceiling in less than a second, even without significant power (yay surgical tubing). The question is... can your driver use it?

-John

[Chris is me]

Originally Posted by JVN

Bingo. We slow ours WAY down to make it useable.

Do you guys slow it down in software, or do you use "extra" gear reduction to achieve this, or both?

[Jstack14]

I find it interesting that it has to be slowed down to be controlled. I guess it depends on the lift type, but suprisingly our lift is rather controlable at its high speed. Could just be the style of lift but maybe look into practiceing some with the arm moving faster (if its slowed through programing). After a little bit you may get the hang of it and have the speed as an advantage!

[JVN]

Originally Posted by Chris is me

Do you guys slow it down in software, or do you use "extra" gear reduction to achieve this, or both?

Software, it is running at about 20% speed most of the time.

-John

[Chris Hibner]

Originally Posted by Jstack14

I find it interesting that it has to be slowed down to be controlled. I guess it depends on the lift type, but suprisingly our lift is rather controlable at its high speed. Could just be the style of lift but maybe look into practiceing some with the arm moving faster (if its slowed through programing). After a little bit you may get the hang of it and have the speed as an advantage!

We didn't have to slow it down to make it controllable. It was very well under control. The problem was the acceleration creates so much force (F = m*a) that some stuff was starting to bend a bit too much and we were worried about the life span if we kept it under that much load all of the time.