Was hoping we could share build season breakdowns, and your success with that schedule.
Kickoff Day: Strategize, Pick what we want the robot to do
Week One: Decide our strategy isn’t going to work
Week Two: Totally veer off of our original designs
Week Three: Find out design is illegal because of a new rule change
Week Four: Consider Quitting FRC/burning down workship
Week Five: Finally let the electrical team wire up the robot
Week Six: Let Software have robot
1 Like
10/10 would recommend schedule
In all seriousness, Team 100 did reasonably well sticking to our schedule with respect to strategy selection. That was a hard lesson we learned in 2017. Prototyping slipped a little in 2018, but not terribly. CAD was where we fell way behind, to the extent that we were essentially doing climber CAD at bag time. This affected fabrication, programming, and electronics, because it’s hard to build things if you don’t have plans. The solution for 2019 will be to have more trained CADers. The hard part about that is not how to drive SolidWorks, but how to impart mechanical intuition and do good design.
We have found clear deadlines and iterations (some teams go up to 7) and having experience CADing specific mechanisms can improve CAD speed.
We took a much more structured approach this year, which helped us execute a much more complex robot than we traditionally build. One thing that helped a lot is that we met on Sunday after kickoff, which helped us space out the strategy and design discussions, so that everyone could read the manual and sleep on the game. However, we don’t meet on any other Sundays. Off the top of my head…
Kickoff Day: RTFM, and decide on general strategy. A little design talk.
Sunday after kickoff: Study robots from similar challenges (2011 for elevators, 2015 for elevators and intakes), decide on overall robot design and start CAD.
Week 1: Design and build drivetrain. Janky intake prototypes (plywood/cardboard/drills).
Week 2: Finish build of drivetrain, wire drivetrain, write basic code for drivetrain, drive around. Begin work on elevator structure. Slightly less janky intake prototypes (aluminum extrusion and actual motors).
Week 3: Battery mount. Elevator first stage structure. Work on intake v1. Test sensors (encoders, gyro) in code, simple auto modes (drive to switch).
Week 4: Elevator first stage powered, elevator second stage structure. Finish first intake revision. Replace plywood electronics board with polycarbonate. Add mechanisms to the software. Test climbs (uses only first stage)
Week 5: Finish second elevator stage, mount intake. Mechanical testing (second stage gearbox was not reduced enough). Practice intaking/placing cubes etc.
Week 6: Lightening robot (had to lose ~3 pounds I believe). Finish the final wiring (make it look nice). Start on second intake revision (withholding allowance).
Week 6.5: Lighten robot, finishing touches. Some auto modes testing, some drive practice.
Overall a pretty good schedule though I wish we could have done more testing throughout the season.
Pre-season: Plan schedule
Kickoff: Strategy
Week 1: Prototype
Week 2-6: CAD, Machine, Assemble
Bag day (6 hours remaining): Wire
Bag day (4 hours remaining): Complete assembly
Bag day (2 hours remaining): Test software
Bag day (15 minutes remaining): Panic because assembly and wire still aren’t 100% complete
Bag day (2 minutes remaining): Bag the bot
Competition practice day: Disassemble, replace parts, reassemble
End of competition: Bag the bot again
Repeat
That’s about how our season went this year. We were hoping for time to fully test the robot and program.
1 Like
Saturday 1/6: Learn game
Monday 1/8: Determine strategy
Tuesday 1/16: Finish CAD
Saturday 1/20: Chassis complete
Saturday 1/27: Robot complete
Tuesday 2/6: Head coach has twins
Tuesday 2/20: Bag it up
Friday 3/23: Setup for hosting district event
Sunday 3/25: Semifinalists, Industrial Design
Saturday 3/31: Semifinalists, Imagery
I’d highly recommend taking a look at 254’s 2014 build blog and JVN’s blog. You can piece together their build season schedules, and it’s interesting to note how they differ.
Like anything in FRC, the ideal build season schedule for a given team is entirely dependent on team resources and priorities and should be designed strategically, just like a robot. They are not plug and play.
The past 2 seasons, we have decided as a team to simplify the design, so we get the robot done really quickly, and have a lot of time to practice. Here was our basic schedule breakdown
Kickoff: Analyze the game
Week 1: Prototype, and CAD the chassis.
Week 2: Manufacture the chassis, and begin designing the intake and arm
Week 3: Assemble the Practice Bot. Start manufacturing some parts for the final bot.
Week 4: Driver tryouts, and testing various intakes (intakes at this point were laser cut wood). Start designing the “ramps” for alliance partners to climb up.
Week 5: Manufacture the rest of the parts for the comp bot, and send parts out for anodizing. Begin assembly of the comp bot. Programmer started working on auto with the practice bot
Week 6: Finish assembling the competition bot just in time for 1741’s pre-bag scrimmage.
Week 2 of comp: Realize the “ramps” won’t work, and remove them before the first match.
Week 5 of comp: Design and mount a climber with our withholding allowance
This season, we ran a switch/vault only robot, but our driver was well practiced, because the practice bot was functional starting week 4. Ended up an alliance captain at all 3 of our events this season (including ranked 2 at our state championship) so this defiantly wasn’t a bad decision.
From what we have found, keeping the robot design and functionality simple still allows you to have a competitive robot. I think this technique is worth considering for teams that don’t have as much experience or resources, but still want to build a competitive bot.
I’ve found in my own experience and from what I have heard from others most competitive teams don’t really have their overall “robot concept” (rough CAD) set in stone until around the end of week 2 and the start of week 3. This only really works if you’ve fully fleshed out prototypes making them easier to transfer in to CAD and easier to program/tune once you’ve completed the robot. By then these teams would of probably started work on the chassis if it’s a year where the chassis doesn’t need to be as integrated into the design (a counterpoint would be 2016, most designs required chassis cut outs and other integrations if they wanted to fit under the low bar). The robot would then be mechanically and electrically complete by week 4-5 with the rest of the build season spent finishing the practice robot, tuning and making any final changes. This can also depend on your competition schedule. On 5406 we left the majority of the work on the buddy climb until week 2-3 of competition season since we knew most of the additions would be fairly minimal and easy to add on in the unbag time.
This is probably the optimal timeline for the best competitive performance and student experience if you have the resources to pull it off. The strategy/prototyping/concept stage is one of the most useful learning tools for the majority of students IMO. Some teams would obviously struggle to pull pf a timeline like this off particularly in years that require a lot of tuning and a more integrated, complex design.
We had limited resources in 2015 and unfortunately we knew that because of the game there wasn’t really any option to reduce the scope of our design strategically like we had in 2014 if we wanted to be successful. The strategic complexity of the game was also far lower than usual. We quickly made the decision that we had to compress the normal 3 week timeline into a single week forcing us to lock down the design and start final cads earlier than we would have normally.
Looking back on it it’s clear that if we had not made that decision we would of had no chance of building the robot we did.
I intend this as a bit of warning for teams looking to copy the schedules of other teams. It’s really important that you organize your schedule strategically for not only the resources you have but also for the complexities of the game. Timeline management really is one of the more challenging parts of FRC.
Our plan
Kickoff day: Brainstorm and figure out how we want to score points
Week 1: Have a fully finished and built Chassis and a fully CAD’ed out design
Week 2: Finish the electrical design and construction on the Chassis
Week 3: Have 50% of the subsystems built and attached
Week 4: Have a working robot and have all the code done
Week 5: Testing of robot
Week 6: Driver practice
We planned on having two identical robots built at the same time so that after we bagged our first robot we could have more driver training and just finishing touches.
How our season actually went
Kickoff day: We brainstormed what we wanted to do, and how we were going to accomplish it. We decided on a claw and a cascade lift to lift the robot up
Week 1: We had an almost fully finished design, and we were running into issues with building the two chassis
Week 2: We had a fully finished design and a chassis skeleton built, but electronics was just started
Week 3: At this point iirc we had to redesign the robot twice already due to build complications, and electrical was almost if not done by now.
Week 4: The full robot design was in its 5th complete iteration by now, we had a claw built and opening and closing, and we had ordered everything that we would need
Week 5: We had most of the subsystems built, but not attached to the robot. (Note: we only built subsystems for one robot and code was working with the second chassis that was just a base w/ no subsystems)
Week 6: We had a few of the subsystems attached to the robot, going into stop build day we still had to attach around half of the subsystems to the robot.
Stop build day: We stayed until 10 at night to finish assembling the robot, we took the claw off so that we could work on it after stop build day, code didn’t touch the robot w/ the subsystems on it at all, all the code was theoretical.
Overall there were a lot of issues we ran into this season. I’m design lead for our team, and the reason for the 5 redesigns of the robot is that we had very limited space and were fighting for every little bit of space we had. I could’ve done much better with relying on my team members more to CAD out things, and I think I was too ambitious with the goal of getting a CAD done by week 1. We were very rushed so rather than try to design a claw from scratch I just took the VEX EDR claw and scaled it up by ~3, then flattened it so it could be plasma cut. Our decision to stick with the claw meant the robot had to be designed and built around the claw, which you can see negatively impacted the structure of our robot.
This year we really learned to be realistic with our goals, and that we can’t be afraid to start from scratch even if something’s already built. I’m also trying to push for a move from a 2 Cim drivetrain to a 4 Cim drivetrain and for more meetings that go longer and start earlier for next year.
Day 1 (Kickoff): Eat pancakes, watch game reveal, spend time being excited because kickoff is exciting, think about reading the rules, start building field elements, build quick & dirty frankenbot using whatever is on hand
Day 2: Actually read the rules, finish building field elements, identify cycles in match play, run timed cycles with quick & dirty frankenbot (frankenbot in testing Sunday evening)
Days 3-14: Iterate systems on frankenbot, explore strategies in depth
Day 15: Run full legth teleop matches with frankenbot and time cycles, strategy and design consensus meeting
Days 16+: CAD (we never did stop)
Day 20: Drivetrains to fabrication
Day 22: Drivetrains fully powder coated, assembled, and running on proto boards
Day 27: Elevators and carriages to fabrication
Day 33: Single claw to fabrication (not confident in design, didn’t want to make two)
Day 37: Elevators and carriages fully powder coated and assembled, claw assembled (still not confident, didn’t powder coat it)
Day 38: First full systems integration, deploy code that had been developed on frankenbot, rerun same timed cycles from Day 15 and compare, decide claw is not the worst thing we’ve ever made but is generally inadequate
Day 43: Final control boards fully powder coated, assembled, and installed; start considering which twin to bag
Day 48: Claw mk2 to fabrication
Day 50: Claw mk2 fully powder coated, assembled, and ready for unbag window
Pre-Season: Make team goal to use more COTS parts, minimize welding. To support this we bought VEX drove in a day frames, VEX DS gearboxes, and a pnuematic rivet gun.
Kick-off: No idea, I was really sick for kick-off. I imagine they ate a lot of donuts though!
Week 1: Decide on 4 wheel drive train so our shiny new gearboxes won’t work on the drive train, also the frame we bought isn’t quite long enough and need to be extended with a peice dubbed “chassis extensions”. I will come to hate the chassis extensions as the build season continues on. Also, the kids kept assembling the “drive in a day” chassis upside down. Note for next season, labeling things up, down, left, right is always useful! We also decided on a continuous elevator with intake claw. Holy moly we need a lot of auton modes!
Week 2 Around this time we decided we wanted to lift two robot at one time, and that we were going to do it with carbon fiber sourced from a mentors work… Oh how I wish I could go back in time. At least those shiny new VEX DS gearboxes will work perfectly for the elevator after doing the math like 50 times just to make sure. Hey the chassis is together, that’s pretty cool! Hey, multi stage elevators you are designing from scratch are kind of difficult to design, not so cool.
Week 3 The elevator is coming along nicely, as is the electrical board! The intake you ask? Eh, we’ll start putting that in CAD next week. Meanwhile we need a very complicated and heavy system on the front of the robot to prevent it from tipping that may or may not be legal all because we want to lift more than one robot!
Week 4 The intake was surprisingly easy to CAD and make, besides the stupid wrist joint geometry that took many days to fit in the robot starting configuration! Meanwhile practice robot is progressing nicely so work has started in earnest on competition bot. Also, paracord was a bad choice for the elevator rigging and the whole elevator makes me nervous because we keep needing to tighten it. We’ll get an update on the ramps next week or something. MAN THIS PNUEMATIC AIR GUN IS DOPE FOR PUTTING TOGETHER OUR CLAW THAT TOTALLY ISNT OVER ENGINEERED!
Week 5: Software wants the robot, Mechanical wants the robot, the drive team wants the robot. Everyone wants the bloody robot. Did I mention earlier that the chassis extensions made life hell? Cuz they do. Comp bot is coming together nicely! The really complicated system to keep the robot stable when climbing now has a part named “the dorrito”. The drivers are doing pretty well and elevator is working great.
Week 6: Comp bot is done! We are scrapping the double lift idea the day before bag up. Yay! Rip the dorrito 
Week 7: Hey 118, that’s a neat buddy climb, mind if we steal it? Cool, thanks! That came together quickly!
Something like that.
Our plan by weeks (worked pretty well in 2017; overextended in 2018 and fell behind)
- analyze game, select/prioritize tasks, design & begin drive train, prototype manipulators. Select field elements to build.
- Build drive train & field elements, continue prototyping, develop architecture designs.
- Complete drive trains & field elements, select architecture, begin detailing designs
- Build primary manipulator(s), install, integrate, test on practice robot.
- Begin drive practice, build secondary manipulator(s), install, integrate, test on practice robot
- Complete competition robot, shakedown practices
- Final measurements, make spares, bag, continue driving
-
- Continue drive practice and improvements with practice robot.
we usually only breakdown after we have a robot.
TBH we did better than last year…
Kickoff: Come up with strategy for Power Up
Week 1: Get frustrated at horizontal cube roller pickups
Week 2: Get stuff CADded
Week 3: Get elevator stuff built and put on practice bot
Week 4: Have functional practice robot. Waste time on ramps.
Week 5: Waste time on ramps. Build competition robot.
Week 6: Waste time on ramps. Oh crap, we have to bag the robot?
If you want to make a schedule, look at your team’s track record and use that to estimate how long it’ll take to properly perform each task*. Consider differences in personnel and resources from past years. Knowing how long other teams spend on various tasks isn’t particularly helpful. 148 gets bent sheet metal parts with very quick turnarounds, but that has little relevance to your schedule.
Do not work backwards from the total amount of time you have. That’s a recipe for inaccurate estimates. Don’t say “we have six weeks, so we can spend one on CAD.” Say “we spent 40 person-hours on CAD for two mechanisms last year, so we need to allocate 20 person-hours per mechanism.”
Once you’re in the execution phase, remember that following the schedule doesn’t really matter, just meeting the deadline. If you can finish something ahead of schedule, that’s more practice time. If you’re behind on a task, you need to shift more resources towards that task or cut scope (reduce remaining workload by reducing requirements) to ensure you meet your deadline.
A typical deadline to shoot for is a functional robot (practice or comp) by the end of week 4.