I feel like one of our team’s biggest area for improvement this year was prototyping effectively and efficiently.
What is the process for teams that prototype well?
What do you construct prototypes out of and how do you effectively test them to help understand how they’d perform in an actual match?
Tricks, tips, suggestions on how you prototype?
1.There are no stupid ideas, even if it is stupid it can lead to the final answer.
2. Have freshman suggest ideas and explain why or why not ideas are good or not. This will lead to better understanding of prototyping and help involve them in the next years of robotics.
3.Sketch prototype on CAD, this year I had a lot of fun using a 2d sketch to make the ball intake work and for the first time we were able to use the cad and never have to make another version because it worked well enough.
4. Contrary to the previous point I made, even if something works see if you can improve on it and iterate many times and expect as many outcome on the field as possible.
5.Use tools like laser cutters or routers to make parts after cad-ing them (only if you have access to them my team doesn’t but i hope that can change soon.)
6. My new idea for next year is to make the field in the first 2 days and have people move the old robots around and see how to minimize motions like turns by making mechanisms like that of 254 (2018) and 1323’s ball mech (2019) I’d hope this would show us where we will spend a lot of time to line up and how to minimize it. I haven’t tried this but I hope it will work.
7. Steal ideas, our intake was an exact copy of 254’s bot in 2016(with large mecanum instead of the small ones).
I hope this can help you protope better
One big thing that I see a lot of teams forget is that the mechanism needs to work on the field in the heat of competition, not just in the shop under perfect conditions. Just because your intake picks up a ball that’s sitting right in front of it doesn’t mean it can get one rolling away from it, off-center, slightly over/under-sized, etc. Test your mechanisms in the worst conditions you expect, then test them again in conditions worse than that. Don’t be satisfied until it meets your expectations in worse than the worst conditions.
For that matter, set requirements before you start prototyping. For example: “our intake needs to be able to pick up a 12-14 inch ball within 5 inches from its centerline within one second”. That way when you’re testing your prototypes you know whether they meet your specs or not. If you find that you’re unable to design a mechanism that meets those specs, you can adjust them but that may mess with your match strategy.
The folks over at RSN put together this video on prototyping a few years back, give it a watch.
It is probably best to design the tests and determine your acceptance thresholds before one designs and prototypes the mechanism. I have seen instances of “wishful thinking” where a sub-optimal concept was accepted after going through testing tailored to make it look good. Either one makes a brutally honest evaluation to determine if a mechanism is good enough or the competition will do it for you.
I did a talk at Champs a couple years back on this - some people say it’s kind of okay. https://www.youtube.com/watch?v=u5lEz4LbiMc
Let me know if you’d like the slides or have any specific questions.
Getting through the process of determining basic mechanism design FAST is a big deal. 4926 struggles with this mightily. So, this summer and fall we are planning a series of prototyping challenges. That means we are going to PRACTICE prototyping with unfamiliar game pieces multiple times. I am selecting “typical” game pieces, dividing the team is 3 or 4 teams, and then giving them 5 hours over 2 consecutive nights to put together a “capture and deliver” mechanism. They wont’ know what the game piece is before I hand it to them.
I’m expecting the 1st one to be really rough, but I hope that as we move along , they will improve in organizing, dividing the work, and executing. I’m really looking forward to it! I think…
If you don’t have a laser cutter or something for fast cutting, 80/20 is good material especially REV’s stuff because all you gotta do is cut the stock to size (even a hacksaw will only take a few minutes) and just tighten the nuts on the brackets. REV has a bunch of brackets that make mounting motors and stuff simple.
Seconding 8020, we order it directly from 80/20 (same principle, just a different brand). Anything that you want to tune (like the distance between two arms of an intake, or compression in a shooter) is worth making in 80/20 to start with, it makes it very easy to make adjustments and dial in on optimal geometry.
The laser cutter was also a godsend for prototyping, once we got geometry that seemed pretty good from our 80/20 experiments we could very quickly make a more rigid and fleshed-out version from wood.
And just another note, I would recommend getting brackets from REV but buying the 80/20 from The Knotts Company. They offer a 40% discount for any FIRST teams, just give your team #.
We really enjoyed our Protopipe system this year.
We have some blog posts from this season showing us using it for our cargo mechanism prototypes.
Thanks for the tip, I’ll look into it!
Here are some build season vlogs that show 1712’s prototypes in various stages (some proof of concepts, others approaching final designs):
I really like this idea. I think we’ll do this too.
That looks really useful! Why did you guys decide to use that over VersaFrame or 80/20?
The Google doc linked on the blog has an explanation for why designed this over using the alternatives at the bottom.
Spectrum Protopipe is awesome and 4926 used it quite a bit for 2019. However, we learned about it from Allen a little bit late and didn’t have time to practice making robust prototypes. Several things we made collapsed and broke under their own weight especially after attaching motors/transmissions. We need to practice; which have planned explicitly as indicated earlier in this thread.
One thing 4926 is working on now is beefing up the Protopipe models to be more robust. Maybe we’ll call it SpectrumPipe! Anyway if we make any changes we think are useful we’ll post them!
When I was a mentor, we depended heavily on LAD: Lumber Aided Design for prototyping, especially of intakes.
Any number of ideas can be tested quickly with 2 x 4’s and other dimension lumber & drywall screws.
An excellent accompaniment were the bearings from Triangle:
https://wpilib.screenstepslive.com/s/currentCS/m/kop/l/1028958-triangle-bearings.
Close tolerances were not necessary because the bearings could rotate in the housings
They often ended up on the competition bot.
A good complement to LAD is CAD (Cardboard Aided Design). We used it to prototype several iterations of the passive Gear mechanism in 2017 in a couple of hours. We used the last version as patterns to cut the polycarb for the final version.