How to Improve for Next Year

Hi, I’m a rookie FRC player and we just came back from our first FRC Competition. It went fairly well but we know we want to improve for next year and we hope we can do this offseason. We don’t have any ‘real’ mentors, the whole thing is up to us(around 4 of us whose this year was the first) so because of this, I’m quite proud of what we did… Regardless, I have a couple of questions for experienced teams. We are mostly looking for places to get help from, like websites, forums, videos etc

-We’ve tried different sensors that we could find but had no real luck due to lack of concrete explanations online or information that we could find… For example, we have a MaxBotix EZ1 ultrasonic sensor that w3e couldn’t get to work because we got all types of info that didn’t match. (if someone could help with that thanks)

-Being the programmer(Java), I feel like I did fairly well but I still have loads to learn. For example, how to test inputs and outputs, how to write data to the smartboard/dashboard, how to use PIDs etc

-Is there any useful mechanisms we should work on that are usually used in games? For example, we saw that most teams used wheeled intake for the cubes and decided to work on that for other years. We also want to work on elevators(to bring the intake up), climbing mechanisms etc. Anything that we can get a good idea of before build season would be great.

Anything else that could help us as a team would be great, we are proud of our work this year but know we can improve and are looking to do everything we can to do so!

Thanks for any Help!

Congrats on the second pick of the number 8 alliance. Can you tell us what happened, it looks like another team subbed in for you in the eliminations?

It also looks like you are signed up for the Western Ontario event in two weeks in London Ont.

So you have time to make changes to your robot in the next two weeks.

Can you tell us more about it? Design, strategy, drive base, how you move the cubes, what sensors you have, where they are mounted, etc? Picture of the robot (The Blue Alliance doesn’t have this years robot).

Hey thanks but what happened is that they picked us thinking that we could do scale, when we actually couldn’t
Because of this, they replaced us lool. Can’t really blame them though…

What happened in the event is that we had an arm and claw, but we didn’t notice the limit on how far it could stick out… We had to cut it and it could no longer reach cubes… We had to be play defense the whole time

As for our robot. We have 4 motors, 4 wheels. A 4 pole frame from which an arm swings from(using a rotating shaft) and from the arm, a spring loaded claw sticks out.
Sensors-none since I couldn’t figure how to get distance sensor working and wasn’t sure how to use the gyro.
Because of this for autonomous we just drove forward, easy but reliable!

For next comp, we want to either change intake or at least lower the frame do that we can grab cubes from the floor…

For next year we want to figure out the distance,gyro, and position sensors(to track movement). We also want to to work on seperate mechanisms to put together to be prepared for all types of robots.
Lastly, which is quite ambitious, I want to try and work with Vision for autonomous.

I know this is is alot, but I’m excited to do better and I think with the right resources and help, we can do alot

Sent from my LG-H873 using Tapatalk

Welcome!

You’re definitely in the right place. Keep reading here all year round and you’ll learn a ton that you can apply to your team robot and strategy next year. You are also lucky to be in an area where you have a lot of local FRC teams that you can reach out to for help.

You mentioned that you didn’t notice the frame perimeter extension rules which forced you to not be able to use your intake. I guess that’s a good place to start for next year. It sounds so obvious but it’s true - as soon as the rules come out, you must sit down and read them, very carefully and very thoroughly. On our team we actually spend a few hours sitting in a room and we take turns reading the entire rules sections out loud. It sounds tedious and boring but it is so important that you catch nuances of rules like this - and learn to read between the lines. If a rule specifically does NOT say something it is probably just as important as when a rule does say something.

Then, after you start building, keep an eye on the Q&A site and the team updates. You MUST stay on top of the changes and clarifications!

You are a nice guy and I like your attitude. Anyways, I wish they had done a good job scouting and learnt your robot capabilities before picking your team.

Many experienced teams have published design documents team 1114, 254 etc. Review their design process and their design, watch some matches with those designs. Try to understand their approach the challenge and their strategy. Also, watch your matches with drive team and see how you could play differently. The view your drive team has during matches is different than the 300’ view. That would be a good start.

Teams 254 and 1114 have their designs on their websites under resources, there are few on TBA. I like Adam Freeman - Team 67, 2012 and 2014 robots and their design document.

Here is a link on TBA:https://blog.thebluealliance.com/category/behind-the-design/

From a build/ mechanical standpoint, I agree with Tungrus. Look at design manuals for teams, such as 254. I always love looking at the small details they implement and it helps with designing our robot, such as things as simple as belt tensioning. Looking at other teams CAD models is a great way to learn also. 973 and 1114 both have a ton of videos on YouTube talking about design methods. Looking around on Chief Delphi is the best way to learn because many teams post the new versions of this information on here.

There was a thread not too long ago with a similar theme and a lot of good overall strategy tips were posted there. Perhaps too late for your coming competitions but should be good food for thought for next year.

Expand your network of connections. Feel free to contact me at any time with questions or needed info on any mechanism/or robot needs. Also another good thing you can do in the offseason as a team is to go through TONS of past year robots from veteran teams to try to understand how they work. Learn what works for what game/game piece and then use proven designs next year. The best way to build a robot is to steal from the best and invent the rest. (watch all of 1678’s workshop videos on their youtube channel, and 1114 videos for CAD)

Just adding a bit of caution, there are too many good teams and lot of material for review and understand. If you see the responses here, there are few teams in common, pick one or two from there and they have sufficient info to read, watch and understand. Then you have to work, practice CAD, build a robot (if you can), if not with metal, go with wood. There will be enough to do. Good luck once again.

To expand a little further about your question about mechanisms: if you look around the pits at your events, and browse the photos and reveal videos from other robots, you will notice a lot of common design patterns. For example to pick up an object from the ground you will often see wheeled intakes, and they’re often in roughly the same configurations. Sets of wheels on the left and right of the object are very popular this year, and most of us are using squishy compliant wheels because of how rigid the game object is. (They came in the kit of parts… so that is a huge hint as well).

It is certainly possible to build a radical new way to pick up the game object. I’ve seem shovels, I’ve seem chopsticks, I’ve even heard about suction cups. But there’s a reason wheels are popular. Go with simple, time tested approaches.

If you study the field of Power Up robots you can see that there are a few general designs that come up over and over. You’ve got arm based pick-and-place. You’ve got forklifts with roller intakes. You’ve got switchbots with conveyor belts. Shooters versus placers. For climbing there are winch based climbers and hook based climbers.

As far as sensors: the best bang for your buck will be to get wheel encoders on your drivetrain. You can do a lot with just that. I recommend after that you buy a NavX gyro. Gyro + encoders are enormous building blocks for your autonomous modes. Vision is cool, but finicky, and I think you will get more bang for your buck if you focus your attention on learning to leverage those first.

When it comes to mechanical design, the best way to improve is by studying other teams. Each robot is the product of an entire team’s experience and ingenuity over the course of six weeks (and usually much more). With each competition bringing together 30+ teams, many of whom are probably much better than you (no offense), they are absolutely fantastic learning opportunities.

Take long walks through the pits at every competition you attend. Look at other teams’ robots, understand what design decisions they made and why. If you see mechanisms/robots you’re interested in, don’t be afraid to ask lots of questions until you understand how it really works- your goal is to understand these mechanisms well enough to replicate them in future seasons if needed.

In my experience, this understanding is usually more valuable than whatever you get out of building practice mechanisms (since the games usually vary pretty widely). A major exception would be the drive train, which a lot of teams like to keep pretty constant when they can.


When you’re new to FRC, your actual performance at competitions is completely secondary to what you learn from them. You might be last seed. You might not get picked. Your robot might not even move. None of that makes a competition a failure. As long as you come out of it more knowledgeable, more experienced, and more confident in your skills and the skills of your team, then you are ensuring improvement and, eventually, “success” by whatever metric you choose. Make sure to have fun too :slight_smile:

All that isn’t even touching the depth of the online resources out there too. JVN from 148 maintains an excellent blog. 254’s technical binders are great resources as well. You mentioned PID control: 971 has lectures on the subject, and this lecture from champs 2015 is really good as well. Lots of top teams maintain really great resources on their websites.

Note: Concerning build blogs, the insight you gain into another team’s way of thinking and approach to tackling an FRC season is usually more valuable than the mechanisms they detail. Though the design details are certainly very helpful too as well


Back to robot design, you can get a lot of information and insight out of videos of other robots and even pictures too. Case in point: back in 2016, our intake was total garbage at our week 1. After it pretty much single-handedly neutered our robot, I took a look at 118’s reveal video and hit “pause” on a shot of their intake. Two days later, we had a working intake.

Accept strategic design as your lord and savior. Rather than describing it, I’ll point you to just one of the great videos on the subject.

Keep iterating. No matter what FIRST says about a “six week build season,” the fact is your robot is never finished. We all tweak our bots over the course of comp season, but don’t be afraid to make major changes and rebuild mechanisms as you deem necessary. Hell, 558 totally (and I mean totally) rebuilt their robots at the Hartford district event for two straight years, and met with great success. Back in 2015, 1678 had to iterate their canburglars several times until they were the best in the world (or best on Einstein anyways. 973’s were pretty sped too). They even dropped a quad-canburglar (!) at one point because it wasn’t fast enough. Iteration is king.

Finally, COTS parts are great, and they’re only getting better (check out all the cool stuff in Vex). Buying things off-the-shelf is a massive time-saver, and you can do a lot with them these days. Top teams often use fancy fabrication equipment (and to be sure that can be discouraging for those without such capabilities), but it’s definitely possible (and I mean possible possible not just technically possible) to build world-class robots with some basic equipment, COTS parts, and a f***ton of hard work.

Given the drive and dedication, I know from experience that it’s possible for a team to vastly improve itself over a single season. From the looks of it, you have a lot more time than that. We’re all looking forward to seeing what you’ll produce in the future :smiley:

Good luck!

All the above is great advice and somewhere I’m sure its mentioned but find a team close to you and meet with them. I assure you they will help in every way possible.

+1 on the encoders
+1 on the NavX Gyros (or pigeon if you have TalonSRX’s)

I haven’t seen anyone mention this, but I’d like to recommend Spectrum 3847’s Design Concepts Presentation if you haven’t seen it already. I really wish I had seen this after my first year.

Also, all of their associated recommended reading here: http://spectrum3847.org/recommendedreading/

In particular, this pointed me to this instructable, which is amazing.

First of all, congratulations for asking the right question! I think this is where most teams go wrong, and the real difference between programs that remain stagnant and keep blowing people away year after year. I think a lot of people above hit some of the actionable things to do. I’m going to present some thinking on how to answer the “How to Improve”.

  1. What went well! You’ll never know where to start unless you know where you’re at. Listing everything out helps a lot.

  2. What went wrong! Doesn’t matter why or how, just list it out.

  3. Look at those pros and cons holistically. Try and see if there were patterns, because there usually are. See if there was any common causality. List all of these out.

  4. Shift gears and figure out what all of you as a team want to achieve next year. Dream big, but realistic. (aka, you’ve never made the playoffs, but next year you’re going to at least the semis or you’ll be a captain at a regional). There’s a 1% you’ll overshoot it, 5% you’ll hit it, but then a really high chance that you’ll end up somewhere pretty great. (sourcing Karthik here). FYI, this is also how Google operates internally.

  5. Once you’ve figured out what y’all want, figure out how to get there! Easiest way to do that is to take the list from step 3 and figure out how to solve those problems over the summer and into the fall.

  6. Use the fall, and hopefully offseason events to run a competition level schedule at 75%. You’ll learn a lot about your team and will give yourself a chance to see where you’re at and how you’ve improved.

  7. Repeat 1-3, and use the season to grow! The key is to give everyone as many chances to iterate and improve. Don’t limit the number of chances you get.

  8. You should do a smaller version of this in between events. That’s where a lot of the best and natural growth happens.

If there’s any "secret’ to 5499’s 2016 -> 2017, it was our ability to do this well. Rooting for y’all to kill it in 2019!

Keep asking lots of questions when you have the opportunity to meet up with teams that are more accomplished. Use such opportunities to probe beneath the surface and ask why they made the decisions they made, What challenges they had to overcome with their design, what is the “secret sauce” that makes their robot work better than the others with the same general features. Do not be a team that only does what the best teams do in a superficial way. Find out why they are better than the rest. Do not be a team that “copies without understanding”. I noticed a pattern developing when interviewing the best of the best Fuel shooters at Houston Champ last year. Make the effort to go to the Ontario Provincial Championship to visit as a spectator if your team does not earn a spot there. Hope to see you there.

Follow Cooper’s advice an take up his offer. He and his fellow mentors/alumni have proven many times on the competition field that they know their stuff.

As usual, the answer to such a broad question is “it depends”. The first step is to do a “lessons learned” session. Figure out what you did well, and what you did poorly. Estimate out how much each thing that you did poorly would have helped if you’d improved it, and how much work it would take to improve it. Start with keeping the things you did well, and working on the “low hanging fruit”, that is, the items with the greatest return on a time/effort investment, and work up over the years. Don’t forget to re-evaluate each year.

And OBTW, try to get some “real” mentors, preferably at least one with design project experience. This will improve not only make each incremental improvement easier, it will likely result in better estimates of impact and cost of improvements. If you can’t recruit an individual in your community, reach out to other teams for help in evaluating your program.

Thanks everyone for all the replies, really appreciate it.
I’ll deffinetely spend some time looking at other robots and teams and see what I can get from it.I won’t hesitate to ask here either!

Sent from my LG-H873 using Tapatalk

Definitely*

Sent from my LG-H873 using Tapatalk

Start SCOUTING!

I personally recommend watching ALOT of robot reveal videos through the years, I’ve delved all the way back to 2008. I find it fun to see what sorts of designs people come up with, and it’s a great way to slowly get a feel for how mechanical teams are supposed to think.
I recommend you start getting comfortable with using JVN calculator to look at gearbox configs, as well as get comfortable with working with chain and sprocket, as that will quickly boost your abilities as a rookie team.