Team 254 - 2014 Build Blog + Tech Binder

FIRST General Forum
#8

Not only do we have a really large team (90 students, at least 30 in the lab every night) but we also have lots of mentors to distribute the workload. I have no idea how the mentors balance robotics with work/family, they are truly amazing.

Also, builds are usually not 8 hours long. Weekday builds start at 5:30 (mentors get off work), we spend an hour eating dinner together, and then most students leave around 10 or 11pm. Weekend builds go from 1pm to about the same time but sometimes end as late as 3am. However, students will sometimes come in shifts for weekend builds.

Our school does not give that many midterms, especially to underclassmen, so we don’t take time off for that.

Despite all that, it does take a tremendous amount of time to be a member of this team, but in the end I think it all pays off!

#9

Until the end of build I don’t think we had a single true break day. We had 3 Mondays that were “breaks” where the majority of the students and some mentors didn’t come, but we still had some mentors and students there for all of them.

We took a few true break days post ship and worked reduced numbers of hours relative to during build, but still quite a lot.

This season was infinitely better than last year. 2013 was a nightmare. It took so much work to integrate all our subsystems and to complete the hanger. We also had fewer mentors, which took a toll on us. There was a 4 day stretch during the last full week of build where I was at the lab from 4 PM until no earlier than 4 AM, with one night until around 6:30 in the morning and another where we didn’t go home. This year we didn’t have a single night that was as bad as any of those.

#10

I’d love to get a link at your Trello board (boards?) from this season. Would you guys be willing to share a link?

#11

We ended up ditching Trello partway through build season (due to difficulty of keeping it in sync) and went with a more analog solution.

#12

Thanks. We had decent success using Trello for our drawing creation, review, release, and mfg workflow. We had about the same experience as you did trying to use it for our overall activities.

#13

This is an excellent resource, thanks for sharing! Now, if only I could get the rest of my team to read this…

#14

Good idea. :rolleyes:

We gave Trello a shot this past year and it was much easier to implement during the fall as there was more time to keep it up to date and time to check it. It almost needs its own staff of a few students to keep it up to date and clear/archive/move around old cards/boards. We found it is a great tool for the offseason when meetings are infrequent and commitment levels are lower.

#15

I’m curious, in the Tech Binder, I’m looking at the Excel Spreadsheets…

For Seconds to Completion and Probability of Completion, are those what you think the “Average” team will do or what you think your team can do?

#16

They are what we guessed/expected would be the average amount of time and probability for the average team in Qualification Matches.

For example, certain teams, especially in eliminations, were able to have a successful catch probability much larger than 10%, but it rarely happened in Qualifications.

#17

To add to what Andrew said, there was significant error in those estimates stemming from the fact that on the day of kickoff we neglected to consider that there would always be up to two opponent robots with nothing to do but play defense.

#18

Gotcha, thanks

#19

The spreadsheet model evolved a lot over time. The version that made the tech binder is close to the “final” version we had (from near the end of build season), before we started to see the game played for real.

It was interesting to watch the model evolve over time. We initially severely underestimated pedestal lighting/ball return time, as well as the ability of the average team to acquire assists fairly quickly by inbounding. As a result, on day one, one robot hurdling, self-catching, and scoring was looking pretty tempting.

#20

Looking over your gearbox data/render in both the blog and the technical binder I was wondering why your team chose to orient the CIM motors the way you did. I’m relatively new to gearbox design and I rather like the triangle method that west coast products and AndyMark use. I just wondering what your team’s reason was for doing it the way you did.

Is it possible that I could get a look at the .STEP file for the gearbox?

#21

The height required to do so would interfere with their mechanisms for ball loading.

It’s a cool variant that can be used to save height, and space on one side of the robot.

#22

The biggest issue was that we couldn’t get the ratios we wanted with the triangular configuration.
Our cluster gear isn’t big enough to get 3 CIMs and a shifting piston around it.

#23

Ah. Sorry for the misinformation then.

#24

What you said was true, also. The ball did hit the cRio and was close to the CIMs when it was fully inside the robot. There is no way it would have fit the other way, ignoring the cluster gear size.

#25

FIRST (:p) of all, thanks for releasing these! The tech binder is just beautifully made! I had some questions though.

  1. It’s mentioned your drive train was designed for sudden bursts of acceleration, but you’re geared for 20 fps+. Why was such a high speed chosen over a lower speed that would have accelerated faster?

  2. You had quite a bit of motors and pistons on the robot, how was your battery life?

3.I recall you experimented with alternate material for bumpers this year, how did they hold up?

  1. No recipes in the build blog? Has the lack of free corndogs doomed us all?
#26

  1. In order to have both high acceleration and top speed, we chose to have 6 CIMs. We didn’t want to sacrifice a high top speed, especially on a field as open as Aerial Assist. Also, one of our initial strategies in Quals was to go for 1-assists with solo-catching, and thus a high top speed would be very advantageous.

  2. The 6 CIMs on the drivetrain were the biggest power draw. The only problems we had were when we tried to push other robots while in low gear, sometimes the breaker would trip. We would obviously replace batteries between every match, and we also purchased some fresh batteries before Champs because the 12 we’d been using for regionals were starting to die.

  3. We didn’t too much experimentation ourselves with bumper material, but we looked at what 971 and other teams did and tried to improve upon it. We used Cordura bumpers for the first 2 regionals and then switched to Sailcloth bumpers after that. The new bumpers had a coefficient of friction as low as .2 (compared to Cordura’s .4 I think). They helped with driving around defense immensely and the material held up pretty well. The biggest problem with the new bumpers was that we ironed on the numbers instead of sewing, and thus sometimes the numbers would bubble up or rub off.

  4. I don’t recall how many times baked things for the lab, he could answer the questions better than myself. However, one night we did make s’mores using a heat gun. It was awesome!

#27

I essentially baked the same things as 2013, in the same order, and I was too lazy to copy the same recipe blog posts over from last year’s blog. :slight_smile: