5119 Open alliance build thread

[Andy-5119]

Week 6-8 Recap

Over the last few weeks, we have been stuck in integration hell, trying to get the robot fully functional for our regional. Unfortunately, this meant I chose to skip the week 6 and 7 updates in favor of taking some time to rest. Luckily, however, now the robot is mostly in its final functional state and I am able to share some details from the past three weeks.

New Super Structure

I love max tube. The structure went together super quickly and was plenty accurate enough for the arm to end up square with the chassis. It is also much lighter than the first iteration.

To mark cuts, we just used this simple jig:

One edge is .25" away from the bosses that fit into the holes in the tube. The other is half inch. By flipping the jig around, we can cut the tube .25" and .5" away from the holes.

New Arm

Everything worked out pretty well here as well. We ended up needing another iteration on the wire management system with a deeper pulley and stronger spring. The combination of these two does a good job of keeping the wire loom from falling off the pulley, an issue we were having initially.

In retrospect, this solution is probably a bit overkill, but it works pretty well, and looks ok. With the wire loom everywhere and the added aluminum angle for running wires, the whole mechanism looks super clean.

Bumpers

This year, we decided to take a different approach to bumpers, in hopes that we can solve some of the issues we have had with them in the past.

First, we cut them on our CNC router, allowing us to add finger joints in the corners. This made assembly easier and made the end result stronger. Most notably, we changed our bumper mounting solution. Now, we are just using studs that come out of the bumpers to clamp some sheet metal brackets mounted on the chassis:

The brackets are just .13" aluminum, what we can get from our sponsor, but they have been holding up great so far. The studs coming out of the bumpers are 1/4-20 bolts passing through t-nuts mounted to the inside face of the bumper plywood.

So far, everything has been holding up great. This solution is simpler and more reliable than our last one, so I am optimistic this can be a setup we can continue to use.

Ballast

The robot ended up significantly under weight, so we were able to add a lot of ballast under the chassis to lower our COG. We got this .25" steel plate cut that just bolts to the bottom of our chassis:

The plate weighs about 30 lbs and brings our COG down to about 11" off the ground, enough to basically eliminate any tipping issues.

Intake

We had some issues with the standoffs that connect to the piston rod ends shearing off. In retrospect, this should have been pretty obvious. To fix this , We added some secondary plates that bridges between a few standoffs that the pistons can now mount to:

The intake has now been fully manufactured out of polycarbonate and has been holding up great.

Some Other Miscellaneous Mechanical Failures and Fixes

Smashed Encoder
The first iteration of our supper structure did not have a hard stop, allowing the arm to smash into the absolute encoder mount if it went past its hard stop. Of course, we had some issues with our code, and ended breaking our only rev through bore encoder in half. This issue has been solved in the latest iteration of our superstructure. Luckly, we did have some SRX mag encoders with through bore housings sitting around, so we just switched to using those. We did not any breakout boards for the encoder pinouts, so I ended up just soldering strait to the ribbon cable, obviously not ideal, but it has been doing okay so far. I think we are just going to monitor the cable for any issues and probably make a few spares for competition.

Arm Backlash
Originally, we ended up with a significant amount of backlash in the arm, the source of which, was mostly the fit between the drive sprocket and its hex shaft. We shimmed the fit with pieces of a Dr Pepper can, and basically eliminated any slop.

Testing Videos

We finally as a chance to get some drive practice in on Thursday and Saturday. A few intaking and scoring videos can be found here

We initially had some issues popping cubes because of a sharp zip tie on the inside of the intake. That is fixed now and everything works as intended. We did not have any issues intaking cubes with the single top roller and a standoff at the bottom to trap the cube. I am not sure why the everybot team chose to use the three roller setup, but this is simpler and seems to work fine.

What is Next?

With our build season coming to a close, we will mostly be doing drive practice and auto development, and finalizing the wiring and other details on the robot. We do have plans to cut some polycarbonate panels that will fit horizontally over the chassis and vertically, behind the air tanks to protect everything and make it look better.

Thanks!

[Andy-5119]

Week 9 recap

I don’t really have much to say for this week. We will be attending the Heartland regional this weekend and the robot is in its final state, so we have really just been doing drive practice. We did install a sponsor panel on the back of the superstructure and some panels that cover the electronics, making the robot look better and keeping the arm from getting tangled inside the chassis.

Drive practice has been going pretty well. We can score about 9 game pieces in our simulated matches with a balance at the end. We also have a handful of simple autos that place a game piece and exit the community.

We are excited to compete at our regional and are hoping that we can end our season with a bit of success. No matter what happens, I am really happy with the robot we were able to put together with our level of resources. We were finally able to build a complete robot that functions at the level we intended.

Good luck at your competitions!

[saatb]

That robot looks great! Looking forward to seeing you all at Heartland this weekend.

[Andy-5119]

2023 Heartland Recap

This year, we were only able to compete at one event, the Heartland regional. We had some last minute fixes on the robot when the bolts holding the shoulder sprocket to the outer arm tube sheered and the motors drove our chain turnbuckle into the sprocket, breaking it as well. We did end up fixing those issues in time, and we arrived to the event with a fully functional robot.

We had a supper rough qualification schedule, so we finished quals ranked 16 despite having the 3rd highest OPR at the event. Overall, the matches went really well. We consistently were able to score two high links every game, and were drafted as the first pick of the 3rd seed alliance along with captain, 5013, and second pick, 1769.

We did end up getting knocked out of elimination matches a little earlier than we hoped. We lost our upper bracket match to the second seeds, 1710 and 1730, and lost our lower bracket match to the fourth seeded alliance. Still, we had a strong alliance and were able to put up some of the highest scores in the event for the matches we did win.

An analysis of our performance suggested that we performed extremely well at our resource level, and that our consistency in scoring suggested we had a robot with quality mechanical design, robust controls, and well practiced driver. That said, we have a few areas the team would like to improve on for next year:

• Better strategic design: there were a number of choices we made that increased the mechanical complexity of the robot, probably without need. For example, if we decided to not use a ground cone intake we could have built a robot with an architecture similar to 7461’s with the pneumatic extension instead of our motor driven one, saving a lot of complexity

• Design for simpler software: the biggest bottleneck for us this season was getting the arm code to work the way we wanted it. We discovered that properly programming our telescoping arm was extremely difficult. I suspect that had we designed a robot that would be simpler to code, even at the sacrifice of some mechanical complexity, we could have better balanced the workload between subteams, increasing our efficiency and performance. For example, using a slanted elevator, similar to what 973, 3512, and 1339 have, might have resulted in a higher part count, but the linear motion would have been way easier on our controls team.

• Attend two events: This one is pretty self explanatory. Attending two events would allow us to learn from the first one and implement changes that would help us at our second regional. Investing more of our team’s resources into fundraising in the offseason could help us afford the extra fee.

From my prospective, we have reached a point in our teams development where we can efficiently design, manufacture and code an effective robot. This year, the robot consistently did what we intended at a pretty high level. We the mechanisms we chose, I struggle to find things we could change that would have improved our overall scoring capabilities. The easiest improvements for our team will come in other areas, like better strategic design and more drive training time (probably outside of build season).

Offseason plans

Along with our normal summer camps and events, we have identified a few other activities that could help us next year:

• Offseason drive training: In the past, we have selected new drivers every year, something that limited their experience in season. Having a driveteam that can work together outside of build season and from year to year could significantly improve their skill.

• Swerve: An analysis of our matches this year shows that our mechanisms were effective, but we struggled getting our cycle times down. Rather than adding custom complexity to our robots, we think that swerve would help us make a huge jump forward in cycle time. I think we could have done significantly better at our event with the same mechanisms on a different drivetrain, maybe enough to win the event. Unfortunately, the biggest problem with serve for us is financial. We have just never really had the funds to spare on buying modules. This leads nicely into the third point

• Fundraising as a whole team project: Our marketing team works extremely hard to pull together funding for us each year, but we think that expanding our fundraising efforts to a wider portion of our team would help balance the workload between subteams and help us piece together more funding.

Overall, we have had an extremely successful season, and our best year yet. We competed with a strong robot, impactful outreach, and a lot of dedication from our members. We have been gaining a lot of momentum since 2020, and with our current group of underclassmen, I expect the team to cary that into the future. Thank you to everyone who enabled us to compete this year and those who helped us along the way! I am graduating, but others will be able to update this thread on our offseason projects. Thank you!