Bad Robots

FIRST General Forum
1

So I apologize if this comes off sounding too judgmental or arrogant, but I have a question I’ve been wondering about for a while now.

My pre-rookie team and I went to a regional this weekend to go see the game played, and check out some of the robots. We were excited for some really innovative, great designs and fun gameplay. But when we got there, we found the reality to be a little different. In the qualifying matches we saw, most of the teams could not even pick up the ball well, let alone shoot, and the brief moments of good gameplay were provided by one good robot essentially playing the game by itself.

I guess my question is multifaceted, first - is this common in FIRST, in Aerial Assist, or was it just this regional in particular. Also, what is the build season like for teams who end up with a robot that just doesn’t play the game well? Is it a lack of resources, a lack of time, or poor strategy? And most importantly, what can my rookie team do, before build season and during, to build a robot we’re proud of and can play the game well?

Again, I’m truly sorry if this came off as offensive, but it was an issue that bugged many of my teammates as well, and we tried to phrase this in the most diplomatic way possible.

2

This is probably just the particular regional you went to, I’ve seen tons of action, but I’m out here in California. As for build season, usually a poor robot is the result of a design that was either too complex and had to be scrapped at the last minute or a design that has been changed too much. Lack of resources is also grounds for a less competitive robot, but I’ve seem some pretty great robots that were made out of garages. To prepare to build your own bot, I suggest you talk about what strategy you’re trying to implement, what you want your robot to do, and then look at some existing designs from others teams for inspiration on how you could build your robot.

3

There are always teams that struggle. Even for “simple games,” building a robot is a difficult task. Some regionals are worse than others. Some years are worse than others. Some teams bite off too much to chew. Some teams don’t have enough guidance. Some teams are impacted by scenarios outside of their control (snow, for instance). There’s no single “magic bullet” to solve this problem, or else it would have been solved a long time ago.

4

And Canada (of course). We still have snow up here.

5

The eliminations would have been far more inspiring. Unfortunately not every team has the resources (human, fiscal, or facilities) available to produce a top caliber robot. The reality of the situation is that building robots is very hard to do, and for many teams a success is just having something out on the field that functions.

6

Well from a team that had trouble last year building a robot (seniors 2 years ago left and did not teach anyone). For us it is really we had trouble trying to figure how to make anything. Robots that don’t work will be in FIRST but its what makes FIRST special because that is when you will see teams that will run up to you guys and love to help you guys. What I have seen that a lack of resources, time and a poor strategy can all be factors in a poor performance.
What we learned this year is that a simple robot that can move can really get you far. We had a basic pickup system that worked 100% and we were able to move. This got us to Semi finals twice. So I guess try to keep it simple and efficient.
I recommend you to get in contact with your local FIRST team and see if you guys can start working together with them. You are also welcome to PM me and the Devastators would love to share all the knowledge we have with you guys. Also the CD is a wonderful place to learn. Trust me I came here last year and I’ve learned so much! Anyone here would love to help you guys.

7

You actually made some smart observations; ones that are pretty profound for a pre-rookie team.

Is this common in FIRST, in Aerial Assist, or was it just this regional in particular

It’s semi-common in FIRST. I don’t know what kind of robots you saw, but at an event of 65 teams, you’ll find that more than 50% of them are “subpar”.

It is also regional dependent. There are regions that have a stronger middle tier (Michigan, Cali, Canada) than others, and there are a lot of factors that play into this.

Also, what is the build season like for teams who end up with a robot that just doesn’t play the game well? Is it a lack of resources, a lack of time, or poor strategy?

You hit the major factors. Some teams have poor strategies, they don’t plan for the resources they have, and they don’t respect the amount of time they have. Usually it’s a combination of the first two that really get teams.

And most importantly, what can my rookie team do, before build season and during, to build a robot we’re proud of and can play the game well?

Before the season: You need to remove as many in-season variables as possible. Of course you don’t know the game, but it’s not difficult to start building simple robots, simple drive trains. Get a handle for the resources you have and how to best utilize them to build effective mechanisms.

During the season: Pick an effective strategy, and build as simple a mechanism as possible to achieve that strategy. Remember, your DT is your base; if it’s not strong, you won’t be strong, and rollers are king.

You’re not being offensive. Some may lash out at you for being “insensitive,” but it’s the realities of the program. Kudos to you for seeing this and taking steps to ensure that your program achieves success.

  • Sunny G.
8

I agree with MrTechCenter

Biggest advice I can give is to not try and do too much in your first year. Too many teams (veterans included) try and do too many things and end up not being able to do any one thing well. Focus on getting a strong working drivetrain If you go with a basic tank drive (also my recommendation), use at least 4 cims). Many teams are looked over in alliance selection if they have a poor drivetrain even if they have functioning other parts of their robot. After you have a working drivetrain, focus on getting one mechanism working for the current years game. Like this year, an arm to posses the ball and let go will make you extremely valuable to teams at regionals. Finally, seek out veteran teams who seem to know what they are doing if you need help (you are already ahead in that aspect). Good Luck!

9

This. This. And THIS.

Robots that can’t play the game more than just defense? All too common. Robots with nonfunctional scoring mechanisms? All too common.
Robots that are too complex? See last two statements…

The three most important parts of an FRC robot are drivetrain, drivetrain, and drivetrain. (And bumpers, which protect the drivetrain.) Keep it simple and tough, then get one thing working in terms of handling the game piece (preferably acquisition) then get one other thing working on that (preferably getting it out of your robot!).

10

It’s fairly common in FIRST, and varies regional to regional. The best advice I can give is to a team that wants to achieve the game goals is to look at Simbotics’s website (simbotics.org) and look over their resources section. Their section on strategic design is invaluable to a startup team.

Also, be sure to look back at some old games and have a few people well versed on what types of mechanisms work well for manipulating certain types of game pieces. Even if the game piece isn’t very similar to a past one, it still gives you some ideas to try to adapt to the new game piece.

11

  1. Is this common in FIRST?
    Its totally event dependent. you wont show up to Waterloo or Orlando finding too many toasters. In contrast, I remember watching the western Canadian regional a few years ago and finding a large portion of robots that could not handle the game piece.

  2. Is this common in Aerial Assist?
    Yes. this years game piece is more inconsistent, bigger, fragile and contested for than almost any other game piece in recent FRC history making it much harder to be an effective scoring robot on the field.

  3. What is the build season like for teams who end up with a robot that just doesn’t play the game well.
    It all depends on when this is realized.
    A. IF REALIZED EARLY:
    Smart teams will change there design to fix the problem. Bad teams will just tell the drive team to deal with it.
    B. IF REALIZED LATE:
    Good teams will try to build a new system that they can switch out at there event or right before stop build day. Bad teams will just ignore/ deny the situation by lowering there standards for what is an acceptable robot.
    C. IF THE TEAM HAS LIMITED RESOURCES:
    Good teams will attempt to use programming/ strategy/ driver skill to compensate for unacceptable or inferior hardware.

  4. What causes teams to be bad?
    Usually a lack of resources or a failure in the design and planning process. If you take 1 thing away from your pre rookie experience it should be think before you act. Try to pull designs or design ideas form past games that used a similar game piece or field layout. A prime example is 67 (hot) and 971 (Spartan Robotics). They both evolved the 2008 world champs robot to make it work in this years game and preform better. 67 is currently 49-7 and 971 is 35-0. I highly suggest looking at some of karthik kanagasabapathy’s recorded presentations on how to preform effective strategic analysis of a game and how to design an effective robot to play it.

Here are some recources:

2014 HOT tech notes by Adam Freeman:
paper: 2014 HOT Tech Notes - CD-Media: Papers - Chief Delphi?

2012 HOT tech notes by Adam Freeman:
paper: Heroes of Tomorrow - 2012 Tech Notes - CD-Media: Papers - Chief Delphi?

Karthik Kanagasabapathy’s 2013 show on design and strategy:

Karthik Kanagasabapathy’s 2011 show on design and strategy:

A CD thread on the weakest/ simplest robot that could make eliminations in 2013:

12

I don’t know who your coach is, but he/she definitely placed you on the best path for your team’s rookie season by attending a regional before you begin.

As a team that has competed in the last two seasons - here is the advice I would slip you to help the process. (It definitely helped us)

  1. Get a hold of your state’s organizers and get them on board immediately with your team. We did this (here in MN we have a great FRC community) and rally them to your cause.

  2. Contact your local Chamber of Commerce/Manufacturers association and ask if your team could present at their upcoming meetings or conferences. Contact your local paper and find a field reporter that would be willing to cover your endeavor. Invite your school’s administration to the meeting(s). Polish the presentation, bring in state reps from FRC, and knock their socks off with stats and figures. And then ask them to help with mentoring, organization, and sponsorship.

  3. Get your ducks in a row early. Starting as soon as possible get the funding in place. Apply for FRC first year sponsorships, organize a few fundraisers to get the cash in-flow moving.

  4. Find a reseller/scrap yard/wholesaler that deals in all things industrial - it is a cheap way to gain structural material (like 8020 components).

  5. Build a skateboard/drive deck/platform as soon as you can afford to and start planning your drive train asap. You most likely won’t have access to 2015’s new RoboRio - so having the CAD plans or a workable deck ready will be paramount. Just keep the FRC notes on pre-season build on hand.

  6. Get the superfluous items off your list early (member sign up, logo, website, shirts, parent organization, etc.). Having this done will free up a lot of time during the build season.

  7. Get in touch with well-established local teams (again, thank you 1816 and 3023) and ask for help.

  8. Attend a pre-regional event.

  9. Set the bar high and expect that your team will have a lot of failures during the build season - but strive for nothing short of the best.

  10. If you need more help, please contact FRC 4607 if you need any other pointers… we are more than willing to help your team where we can - so many have done the same for us!

GOOD LUCK!

13

I will add one thing to “Chief Hedgehog’s” great list: Set yourself the goal to complete your build in 30 days rather than 45…

(1) If you succeed you will have a solid two weeks to give your drivers practice time - you can never have too much practice time…or…
(2) You will have time to build a second robot for competition and use the first one to practice.

(3) If you fall a bit short, you still have time to finish

(4) You get the opportunity to see what parts of the robot don’t work quite the way you planned, what items come out of adjustment easily, and what things require constant maintenance.

(5) This gives you the opportunity to show your robot off to your sponsors and community to get them excited and increases your financial viability for the following year.

As to the observation about the number of “bad robots” I tend to agree, and (without invoking a great deal of angry replies) I believe most of the “reasons” that are cited are just excuses. I inspected during week one in Palmetto, and was astounded by the number of teams who came with a robot that was far from finished, or marginally useful at best. The Andymark AM14U drivetrain was an awesome resource this year, which meant that any team could have a great driving base before they left kickoff (do all kickoff events have a robot quick-build session?). Even if that is all a team managed, they would have been useful to an alliance this year. Especially if they spent the rest of the build season practicing and devising strategies.

14

Most of these pitfalls can be addressed with understanding.
Before the season: Understand what your manufacturing, financial, and physical capabilities are (Possibly do a SWOT analysis). Address any weaknesses, or understand them.

During the season: Understand the challenge. Don’t start off by ‘building something’, instead start off by breaking down the challenge and attacking those portions of the challenge that your team can handle. Set milestones and put expected dates to those milestones so that you KNOW where you are in your schedule.

Planning and an understanding of what you are capable of (and, more importantly, not capable of) will avoid most issues that create a ‘bad robot’

BTW, your post is neither Judgmental nor offensive. It sounds like you are trying to avoid the pitfalls others have fallen into. That’s what Chief Delphi is for.

Good luck to you.

15

I see your FTC team is based in the Capital Region, so you must have either attended the Greater DC regional or the Chesapeake regional.

In general, this region was especially hit hard this build season with close to 10 snow days and school closures. This is a significant factor when last year we had at most 1 snow day. So you can see how this drastically lowered the competition.

Combine that with the Capital region being a middle of the pack region when compared to competitiveness. We also have only 2 or so regional powerhouse teams, so we do not many extremely strong teams unless national powerhouse teams come to visit the region. (233 visited Chesapeake this year, 192 in 2012, 365, 79, etc). As a result, the competitions are normally middle of the pack, DC a little below average and Chesapeake a little above DC.

Hopefully moving to a district system will help improve competitiveness next year, however I feel that the competitive will actually go down next year. The capital region has always had a large amounts of MAR teams come down to boost our competitiveness, and districts will eliminate a lot of these teams.

So in short, you saw a middle of the pack regional which had an especially weak year. Other years, Chesapeake and DC are more competitive, but not by much.

16

So, this is important, yes, but I’d be careful not to go too far onto this side and forget to challenge yourself. I’ve seen many a team and leaders who have preferred to be complacent and not attempted to try to go the extra mile.

There’s a line between ‘not doing too much’ and betraying the program by not challenging yourself. Toeing the aforementioned line is what teams of all levels struggle with.

  • Sunny G.
17

Understanding the logic behind the Minimum Competitive Concept in conjunction with getting early practice building drivetrains in the off season (and driving the heck out of them) will put you off to a great start. Depending on your region, there are probably veteran teams in your area with old kit chassis that either went unused or are obsolete. Be sure to contact veteran teams in your area to see if you can have one (special shout out to 360 for doing just that). For a control system, even something as simple as a 4 channel RC airplane transmitter/reciever combo and some spare motor controllers (jaguars, talons, victors, etc.) is sufficient to replace a $500 cRIO/digital sidecar combo if your only goal is to drive around (though you’ll still have to get a battery and an electrical distribution setup going). Our first “robot” did nothing other than trundle around at a very slow clip, but if you set up a basic obstacle course, you instantly have a Tank Drive 101 class for new drivers.

We used essentially this package right here. It won’t get you programing practice, but it’s a very low-cost start.

18

All of this is very true. My team has made better robots in our last two years than we have ever made before, and it is because we have done specific things very well. In 2013, we focused on getting a really consistent shot. We did it, and ended up with our first regional win ever(albeit as a second pick defensive bot). One of the first things we look at is drivetrains. In fact, it is a deciding factor in who even gets on our pick list and who doesn’t. Very, VERY few weak drivetrains ever make it onto our picklists. A good, powerful tank drivetrain means that even if your bot isn’t the best at playing offense, you may get picked for your strong defensive capabilities. Look at 5125 at midwest regional. Their entire robot was a simple kitbot drivetrain. However, it was done so well it could push almost every robot at the event. Even through there were still collecting and scoring robots still available, they were the first alliance’s second pick and ended up winning the regional.

To answer OP’s questions, it really does vary based on what regional you attend. There are some locations where most teams have great resources and can make great robots. Other locations may not have as many resources, possibly caused by a lack of students, mentorship, or sponsorship.

19

I DEFINITELY agree with hedgehog on the failure aspect. One of the things he didn’t mention was the the main feature of our robot, namely a large shield, was actually making the best out of a bad situation. we had meant to make the robot climb the pyramid (the endgame last year) and get 30 points. after we failed at doing that, we decided to build a shield, making use of the existing mechanisms. It worked for us. (it also made a stylish way to show off our tshirt)

also, a corollary to failing, everything will break. We literally had no system on the robot NOT break by the end of last year, the omniwheels even fell apart, and the kit wheels are BALD (to the point where i’m surprised the DMV didn’t say anything :wink: )

Good luck next year, I hope you do well! You’re already thinking like a veteran. just be prepared to work some late nights, and NEVER lose hope in yourself, your robot, and most importantly, your team!

20

Congrats you have experienced the great divide.