Failing in 2021 and what i learned

looking at the numbers my team (5414) had an amazing season, from chairman district champs to IR@H finalist, WFFA, and the designers award for the game design challenge. The entire team including myself is extremely happy with how our season went. This year has taught me so much and it wasn’t primarily because of our success. A large portion is what I learned was on the hidden failures that are unknown but to a few outside observers.

Fail Early and Fail Often

While we’ve never bagged a competition ready robot by the time we get to our first match we can at least drive, but this is the first year we have had a robot not finished enough to compete, and that is such incredibly scary for me to admit. Everyone talks about how we should fail early and fail often so we can learn quicker and while there is some merit to that I think there’s another reason that few people talk about.

Failure hurts, and nothing I do or say, nothing you do or say can change that, it’s simple human nature. Even now, weeks after the competition has ended, I can still feel the sting of this failure, and it is going to be something I have to co-exist with for the rest of my time on 5414.

Ed Catmull, who is the president of Pixar animations has this great example in his book creativity inc., when learning to ride a bike you should buy the bike that sits the lowest to the ground and wear a helmet and pads. Because when you fall off, and you will fall off it will hurt a lot less than if you were learning on one of those comically tall bikes without a helmet. The reason you should fail early is because you fall a lot shorter than if you design and manufacture a robot only for it to not work in the end.

The Meta is not Always Right.

In 2020 I didn’t pay any attention to chief Delphi or similar places where FRC students and mentors can gather and discuss design. Over the pandemic I have become increasingly dependent on these sources for my ideas and on kick off I made the mistake of having these sources open on my second monitor while I was setting goals with my team, and this led me to get sucked into the Meta.

I can say beyond a shadow of a doubt that if we had followed the meta of a swerve mini bot with just an intake and no shooter we would not have done as well as we did. In our group we had the highest accuracy and power port scores which allowed us to make it to finalist and I believe no matter how much went right for us we would not have been able to beat team 359 who was the winner of our group if we followed the meta.

This has led me to realize that I cannot rely on other people for goals or robot ideas for my team because the meta might not be what is best for the team. This has led me to make a promise to myself and my team that the first 3 days after kicking off I won’t be on any FRC related forums.

Communication is Key.

After the season was over, we began an intense effort to overhaul or shop’s organization and infrastructure. As one of the leaders of this effort I reached out to some leadership members to see if they wanted to come in and organize their respective areas. When our safety captain came into organize our first aid supplies, he walked past the table where our 2021 bot’s parts where still strewn across the tabletop because I couldn’t and still can’t work up the courage to clean that table.

As it would turn out he had no idea we were building a robot, something I thought was obvious, we are a robotics team when everything is said and done. But of course, it was obvious to me that we would build a new robot, because I was there when we made the decision. Meanwhile the game design people or the people who were not at the kickoff meeting had no idea. This revelation led me to connect the dots on a few other reasons we failed that I will discuss in a later section.

Returning to Normality is Hard.

The last in person meeting as a team before the pandemic struck was our week two competition at Plano. We would continue to meet over zoom and talk over slack, but we saw a huge dip in engagement and attendance as many teams did, especially once we knew that 2021 was going to be “at home” challenges. Not only did we take on less rookies than we have been the last two years we also saw a sharp drop off of our veterans and even a lot of the team leadership became unengaged.

Looking back at it I must give massive credit to our lead mentor Andrew (who is the WFFA for FiT this year), this season has given me perspective on what a challenge it was for him to try move the team completely virtual.

When we were out lining how we wanted to approach the design of our 2021 robot the lack of engagement we had seen from the team played a factor in the discussions, but we had assumed (wrongly) that once we began meeting in person again all the people who had been radio silent would magically re-appear.

Looking back that was a bad expectation, this year was hard on myself and my fellow students and many were too bogged down to even think about the possibility of robotics.

Burn Out

Burn out, like failure, is a much-discussed subject in the FRC sphere. This year has given me the first taste of what burn out feels like. Nothing could have prepared me for it, most people talk about how to recognize the signs, having read those articles and sat thought those talks I thought I could prevent myself from burning out. I was wrong, by the time I realized what was happening it was already to late.

Most of the design was completed by me and one other person, with support from some others on minor things. As you can imagine that put a lot of stress on us and I would routinely find myself cadding 4 hours every day of the week, which is not an issue of itself. I regularly compete in cadathons in which you design an entire robot in just one week. But when this dragged on for 4 whole weeks it starts to wear you down, we wanted to have the cad done on the 30th in time for our first meeting but as that deadline approached found myself unknowingly making up excuses not to cad. Sometimes I would cad and just sit there staring at the main assembly for hours on end not doing anything without realizing it. Looking back at it that is textbook definition burn out. But at the time I didn’t know I was burned out, at the same time the other person working on the cad was unknowingly burned out as well and this brought the entire robot design to a virtual standstill. This made us miss the deadline by almost 2 weeks and we were rushing to try to keep up on our manufacturing pace so we could get this done on time. This Manufacturing problem was compounded by the fact that a lot of teammates either unable to come in, or similarly struggling to find the motivation, the manufacturing queue stacked up quickly.

Putting your Eggs in One Basket

From the offset of the season, we knew that maximizing our in-person time was of up most priority. We are lucky enough to have a CNC router and we wanted to maximize its usage for the fastest possible manufacturing. In plain terms we put all our eggs in one basket, and we ran into an endless number of problems when trying to CNC something. Part of this was caused by our lack of CNC experience and the fact that the people who did the CAM and ran the CNC where completely detached from the design team, which is something I hope to address in the offseason. It took us until we had two weeks’ worth of back log that using just the CNC was not going to work and so we began a frantic effort to get someone on the mill so they could manufacture many of the tubes we needed.

Reflection

Reflection is important when you fail, lots of times we do it without realizing; and While a large majority of reflection is done in our heads. I’ve found that the power of writing it all down and writing this out like it’s for a stranger has helped my reflection on this season an order of magnitude more than if I had just kept it rattling up there in my brain. Reflection is something we all should take the time to do, and it’s not just something you need or should do alone. Many of these points I’ve gone over with the people who were directly involved in our robot this year to help my fellow teammates, because I know there just as sad and disappointed as I am this robot never got finished.

The Light at the End of the Tunnel

After talking about failure, you have to discuss your success and while we had a lot of failure we had just as much success. Our 2021 robot was a swerve and while it’s not completed after the texas cup we’re going to get the base up and running so we can have a swerve test bed for the programmers to experiment with because the new control system makes swerve more viable than years before for our team. Additionally, it made us cut down and optimize what we spent our in-person time on while they were not all good there is a lot of practices, we adopted this year that we could carry on to the next years. 2021 also allowed us to explore the concept of using gas springs in place of pneumatic pistons for intake deploy which I think could be very valuable for future designs. We also learned how badly the disorganization of our storage could hurt us, looking back we were ordering stuff and waiting a week for it to ship, when we had it in the shop but couldn’t find it.

TL:DR

I’m approaching 2,000 words so I’m just going to re-state my points here. If there is one thing that I want people to take away from this is that failure hurts and trying to keep it from hurting is wasted effort. I’ve learned that while the views and ideas of other people outside of my team are important, I need to stop letting them dictate to me what I should do all the time. I learned on a team as large as 5414 not everyone is going to know what seems obvious to me or the people in my subsystem. I need to realize that returning to “normal” or whatever the world looks like post covid is going to take a lot of time and hard work and I shouldn’t expect my team mates to come back magically. After my encounter with burn out I think I’m more equipped to deal with it in future and help protect my teammates from it. I’ve learned that putting all your faith in one or two people or in one or two machines is going to drive those people and those tools break down much faster than they would otherwise and expanding the number of people and machines involved can vastly increase creativity and productivity. Most importantly (to me at least) is the new appreciation and importance I have on reflection, being able to looks back and see what went wrong, and while hindsight is defiantly not 20/20 it’s a lot better than being blind.

2021 was still a great year for pearadox despite what I said, and with the things myself and my teammates learned we’re ready to tackle whatever the 2022 season may have in store for us head on.

61 Likes

Good post @beter, and as a mentor it’s great to see the growth in the students on the team. Sometimes I wish I could teach these lessons so students don’t have to “learn it the hard way”, but that can be hard to do.

I think this is an important lesson and relates back to one of Karthiks golden rules of “build within the teams limits”. Sometimes going with the meta isn’t building within the teams limits, or maybe in this case, utilizing the teams strengths.

In talking with @Allison_K, she also made a very good point that in here is recognizing that a season can be a success even when it doesn’t feel like it is. It’s completely OK to be able to look at a list of accomplishments from the year and struggle with knowing where you wanted to do better. Conversely, it’s important to be able to feel like there were significant “failures”, but to be able to look back and find and acknowledge the successes.

I’m looking forward to seeing where our team is able to go this summer and fall, I think we’ve got a bright 2022 ahead of us :grinning:

22 Likes

This is a mature and well thought out reflection, good on you for posting it and sharing, even when it’s not all sunshine and roses.

Conversely, it’s important to be able to feel like there were significant “failures”, but to be able to look back and find and acknowledge the successes.

Your mentor Andrew pointed out something you’ve really shown to understand, especially in seeing you work on Discord as your processes visibly changed. Keep up the work, I can’t wait to see what 5414 does next.

7 Likes

Beter: Great post, was a good read.

This. The meta can get your average team maybe a little better but overall it is harmful and leaves you floundering when alone. There is obviously an FRC meta on how to build your robot and run your team, and many in the “hive mind” strongly believe this and are rather annoying towards people who they talk to that don’t follow the meta. This is the origin of the “Big FRC” joke, an overarching organization that controls the narrative in FIRST and is essentially the puppet master to lower the playing field of competitiveness. The embodiment of the FRC hivemind. At least that is what I was thinking when I first started mentioning it.

In 2019, the early season meta/widespread idea at lower levels of communication was velcro for hatches. This was a terrible idea and most people realized this by week 1.

In 2020, it was vectored intake wheels. They need a relatively specific set up to be done well and most people reading this meta would blindly copy some numbers and think they’ve done something right when in reality these didn’t work well. You needed to understand where ViWs work well, what they need, etc. to make it an effective mechanism, something you can get from prototyping and testing during the season.

Organic and creative idea building is what sets you apart in terms of your robot design. That isn’t to say that learning from others isn’t important but critically thinking about what you learn and take from others is a skill that not only applies in design, but in life, politics, and more. Looking at what other teams are doing in a season is important but you can’t just look at what they are doing and try to copy it, it won’t work. I feel like I could’ve fleshed this out better and more coherently but alas class.

Not looking on Chief for kickoff week (well for robot ideas, there are sometimes important updates/arguments on there), not looking on discord, etc. is important. Throw ideas off your team, prototype, and then look to others. Eventually you get tunnel vision.

I want to echo what @krf said, this is well written and inspiring, thank you.

7 Likes

Stuff like this makes me think that FIRST is just as much “for” creating a mature sense of self-awareness as it’s “for” creating inspiration and recognition of science and technology.

All-- this is a person who knows what water is.

15 Likes

Excellent introspective! Being able to do this and synthesize down to discrete themes and thoughts is very advanced self/team reflection. Kudos to you and 5414.

I’m wondering if you can expand your thinking a bit more on the above section. What about your reflection of IR@H makes you so sure that mini-swerve would have been the wrong choice? Are there specific elements of that type of robot that you think would limit you? Obviously 5414 did great in the challenges so I think you guys made the right choices, but I think its super interesting to dive into this type of thinking. Can you use your results as justification for not taking another path?

Congrats to 5414 and particularly to @ahartnet on a great year :slight_smile:
-Brando

2 Likes

When i say that the meta this year was not for us I’m not saying that we wouldn’t have been “successful”. In fact our 2021 design was a swerve mini bot although it did have a shooter, but if we removed that shooter and just did what teams like 148 or 359 did we would not have placed as well because at the end of the day these teams have many more resources then we do and while we could do a swerve mini bot with just an intake “successfully”. It would almost certainly not preform as well as those teams and we would lose the advantage of being able to shoot which is what ended up pushing us into the finalist position in our group.

1 Like

To expand on what Beter said - the resources we were limited in is prior experience with swerve, time to manufacture parts, ability to do any software development/manufacturing outside of our shop (which was limited to 1 day a week until the final week), and number of students actively working on it (which was primarily 2 people doing design and manufacturing, and 2 people doing software).

When I did my own reflecting I wonder if we would have been better off I told the students that due to time restrictions we would only use the 2020 bot from the get go. With the extra time spent refining our 2020 robot, I wonder how much more we could have cut down on our Autonav times, hyper drive times, and completed the galactic search. As it is, all of our autonav times are from 1 long evening of path development and testing, and about 4 days of practicing the hyper drive. We’d also be better set for the Texas Cup

But I think the students obviously learned a lot and I think the long term value from that is greater than even if we could have been winner vs finalist. We barely eeked out finalist though, and I’m not sure the 2020 robot would have been capable of fast enough times for autonav/hyperdrive/galactic search so all-in-all we may have had the best outcome we could have hoped for, haha.

3 Likes

Thanks Andrew and @beter for sharing! I think many FRC team struggle with this type of decision making, so having an example to openly discuss is great.

125 made a lot of similar considerations, but with slightly different restrictions we arrived at a different conclusion. Using the 2020 robot would have been the most straightforward for us to do specifically for IR@H challenges, however we had 0 lab access, and any tuning and video taping would likely have to be done outside (in New England…in the winter).

We ended up deciding to finish and execute a swerve design we’d been poking at off/on for a few years. We figured this was more executable at individual homes, with the various pieces required flowing in from suppliers, printers and a one-time lab access permission to make some plates. We were fortunate to gain lab access a handful of times in the last 2 weeks of the skills challenges and scrambled to get up and running and record our times. Had we not gotten that lab access, we were planning on running the robot on the blacktop outside.

Because we were focused more on future development, we didn’t miniaturize our swerve, we just made one that could be a good base for the eventual transplant of our 2020 robot on top. Reflecting on the choices we made, I think we made the right call by not considering the shooting challenges at all - they just require so much time and access we simply had no line of sight to. The choices after that? Probably could have made some better decisions on the actual robot we executed to be even more competitive, but that is always the beauty of hindsight. We’re plenty happy and grateful to achieve the results we did though.

Thanks for starting the thread!

3 Likes

You hit the nail on the head. When I first heard “failure leads to learning,” I refused to accept it because in my head, the only type of failure was riding the ridiculously tall bike with no helmet - failing a class or not getting into XYZ college. It took me a long time to realize that good kind of failure is not getting a homework problem correct and asking for help to fix it - riding a smaller bike.


This is so true for many things but especially burn-out and it’s cousin, depression. I don’t have much to add here other than I’m glad you mentioned it - FRC is a perfect storm for students and mentors experiencing burnout. Can we come up with ways to recognize burnout in our team members? What do we do if we see it happening? @P.J might be a great resource!


Absolutely! I maintained a build blog for a few years - I found real value in being able to write down what happened that day(s) and reflecting on it - it allowed me time to assess what happened and plan instead of just react. We don’t have to wait until the end to reflect, you’ve pulled so many great insights - a semi-weekly or weekly reflection is a great way to slow down and take a step back.

5 Likes

In hindsight I wish we would have built a new robot this year, we just didn’t have the time or the resources. When you think about playing these challenges with a robot built for a full game, it is so much more than you want to even think about. The biggest one is time, not even in the build but the hours and hours of trying to shave a half second just to have a smaller faster robot shave 3 full seconds.

It sounds like you all made the right choice for your team. Trying to dive a Cadillac through a go kart course was fun, but incredibly stressful and obnoxious. I think it sounds like, even though you had challenges, your analysis was spot on.

2 Likes

An excellent description. It very rapidly became clear that our 2020 “Robot De Ville” was not going to be competitive on those challenges. Had we had full access to our usual facilities and students we might well have gone down the purpose-built for 2021 tiny swervebot road. In the end, I’m glad that our students chose not to. This year we at last had the opportunity to really iterate on our mechanisms and our programming—even ones irrelevant to the 2021 game like our climber. And I think that will turn out to be more valuable to our kids and to us as a team in future years than blue banners.

2 Likes

We knew going into the 2021 season that not being able to build a custom drivetrain centered around the 2021 at home challenges, we were going to be at a disadvantage to teams that had the ability to do so. That’s ok… as usual in FRC (and especially this year), winning isn’t really the end goal.

To OP: Absolutely awesome write up. Thank you for sharing your experiences - it takes time and maturity to rip through something like that. It resonates well with my experience this year, and what I anticipate going forward. Thank you again!!!

4 Likes