19564 ITKAN Solar | Behind the Bot | FTC DECODE Robot

[00:00:02][music] >> Hello everyone, reporting today for Fun Robotics Network. I'm Carson [music] and with me here today are Omar, Jonah, and Abdul from team 19564, It Can Solar.

[00:00:15]They were the winning alliance first pick at the Jackson division at your Houston World Championship. Currently ranked second, undefeated at the Multi-National Tech Invitational. This one of the most specialized close zone robots in this DECO season. I can't wait to jump into all of the hardware, fantastic iteration process, and really, really simple but very effective software coming up on this behind the bot.

[00:00:37]All right, guys. So, first question has to be about your iteration process. You guys have so, so many different robots.

[00:00:42]Just tell me first, how many robots have you built this season?

[00:00:44]>> Yeah. So, very unique part and is our iteration process. And so, we've actually built 14 fully independent movable robots. So, that's 14 robots that we could individually drive and we could probably make a full line with them if we have them all here today. And so, we do this by our very, very like quick iteration process. We use this with the simplicity of our robots. So, for example, we have we're using just base channels on most of our pieces. And it's a singular plate on the outside, singular 3D print holding in our wheels, and then two very simple inner plates with just directly driven and then uh gear or uh sprockets on the transfer and a directly driven shooter with a static hood just to keep By keeping the robot simple, it means that we can um iterate it on a lot. So, like before worlds, we were iterating like um 50 hoods, I think, just because we could very easily print it and just put it on with the meaning like six screws. And I think at our peak, when since the robots are so simple and so fast to build, we were able to just like, I think in around a week, we were able to get three fully built robots out. This helps us prototype a lot of our uh mechanisms, especially our like the diverter and our shooter, which require a lot of work.

[00:01:49]>> Yeah, talking specifically about materials, I know you guys are using that SRRP this season. Have you been using it for a couple season? What is your advice to teams uh who want to use this material? Any tips or tricks you have regarding it?

[00:02:01]>> Yeah, so this is our first season that we used the Swift SR PP material. And so, we actually found a a lot of good success with it because before we would use like Delrin or like acetal, and we found that that was a little bit more brittle and also a bit heavier. And so, we could actually by pocketing the SR PP and using a lot of channels, we have around like six or like five or six channels on our robot, that allowed us to be able to have a high rigidity even though SR PP might be a bit more flexible. Um we can also it also allows us to pocket it, so we can have incredibly lightweight robot like around 20 lb for everything. And that would basically just allow us to uh have this material that's super light, but also very easy to cut. So, we cut all of this material in house. So, most of our robot is just 3D printing and SR PP, which we can uh cut [music] in house very quickly and get like full robots out in in just a day.

[00:02:47]>> Yeah, talking about the thickness, is it like 3 mm, 8 in, what thickness, 2 mm?

[00:02:51]What is this SR PP? Do you use different thicknesses or everything's the same?

[00:02:55]>> Yeah, so we actually use 3 mm. We haven't really like gone off of that.

[00:02:58]We've used 6 mm before, but we just found that it be it's not really necessary for FTC. We don't need that level of rigidity, especially with the amount of channels we have. We can keep that level of rigidity with just the 3 mm pocketing, like very very heavy pocket.

[00:03:11]>> Before we get to our next segment, we'd like to thank the following.

[00:03:14]Take on the Decode season with AndyMark.

[00:03:16]FTC teams can discover great components such as AndyMark's 3-in mecanum wheels, programmable servos, sensors that detect distance, color, orientation, and many more solutions for your team. Find this and more at andymark.com and count on AndyMark for the reliable service that teams expect.

[00:03:33]>> Yeah, let's let's jump into the drivetrain. If you guys can tilt your robot up, we'll take a look at the bottom now. Uh I see you guys are running some sort of Swift wheels. Are those custom? What are they? Talk me through them.

[00:03:42]>> So, we are using the 110-mm Swift drives, and then we use custom TPE rollers that we uh self-mold. And uh from there, we'll be swap them out with the uh default and the or the cot rollers. And this gives us like a very very good grip so we get a lot of speed.

[00:03:57]Right now they're very worn out and we plan to swap them like tonight. But for right now they work very very well. Like when they were very fresh at the world championship it gave us immense speed.

[00:04:06]And using the larger radius wheels also gave us a bit more speed.

[00:04:10]>> Yeah, talking about like the iterations, you know, 14 iterations. How did the drivetrain change through this if at all?

[00:04:15]>> So, initially we started off with like a more square robot as people might know from the beginning of the season. And from there we kind of kept the same drive. We actually started off with a smaller Swift wheels and from there we just ran with those in the drivetrain.

[00:04:27]It's like pretty simple. We do a very unique drivetrain design where we don't use parallel [music] plates or it's like a version of that where we have an outer plate and we have a 3D print that's only clamping here opening up space in the back like how you can see this axle and motor are just able to sit freely. And that basically we run through that design the entire season. The main change that we made was just making it more more robust. So we make the 3D print one full continuous piece before between the two. And when we switched it out of the diverter in our robot, we basically just switched it so that the wheels moved closer together and around like V10 we switched to the larger wheels just for worlds and like regionals.

[00:05:02]>> Before we get to our next segment, we'd like to thank the following.

[00:05:05]FRC Teams' managed store program is an easy and profitable way to fundraise for your team or off-season event. FRC Teams will handle everything from production, packaging, bulk or drop shipping, payment collection, and accounting.

[00:05:17]Teams can set their own prices and FRC Teams will take care of the rest. Fill out a quote request or get more information when you schedule a callback at frcteams.com.

[00:05:27]>> Yeah, let's talk about that diverter.

[00:05:28]Now this is easily one of the most unique parts of your robot. It is so so effective. So first if you guys could just demonstrate how it works to me and then I'll ask more specific questions.

[00:05:38]>> So yeah, basically the diverter allows balls to come into here and get pushed in. So let's we can do that one more time. It's an outtake. So the balls come in here and just gets instantly pushed in by the rollers. So, these wheels are actually very very incredibly specific spacing. Even like all the cuts and everything have been tested for like all night just to see get the perfect like level of compression. So, that way when balls fall into Auton, they get pushed down and not like moved out in any weird way.

[00:06:03]>> Yeah, I want to talk about these wheels specifically. Did you guys cast them yourselves or are they purchased? What are they?

[00:06:08]>> Yeah.

[00:06:09]Yeah, so these wheels are actually just the 4.5 or the 3.5 mm Swift wheels, but we actually just cut out these centers over here so that they would become star wheels [music] cuz we found that would be more compliant to the balls falling out of the game.

[00:06:20]>> I see. And talking about the powering for this, the diverter is running incredibly quickly, but it's just off of a servo. What is that gear ratio you guys have running on that servo?

[00:06:28]>> Yeah, I think right now it's about uh It's about an 8:1 right now.

[00:06:32]>> 8:1, okay.

[00:06:33]>> But we just wanted to maximize our servo use because we use a lot of motors throughout the robot. So, we found it to a lot more helpful if we could save one for this unique mechanism.

[00:06:41]>> I see. And as far as like switching the diverter, I know you every time between matches if you switch alliances, you have to switch sides. Uh how fast [music] is that process? What advice do you have for teams that want to make modular systems like you guys have?

[00:06:52]>> Yeah, so my advice to teams would be would to make their mechanisms as hot swappable as possible. You could see here on our robot, all you have to do to take off these diverters is these these three screws on both sides. There's four screws in the middle of this channel and then three screws under. So, we were able to switch it out even during the Vinci matches where there's back-to-back matches uh that are rapidly happening with just little amounts of uh switchable parts. And then also for the wiring section, we have two different wiring strands on each side of the robot so that whenever you switch it in, you can just plug in that motor and you're ready to go.

[00:07:23]>> Awesome. Talking a little bit about sensors, I see you guys actually have a dead wheel or something at the bottom of your diverter. What's going on there and what is it used for?

[00:07:31]>> Oh yeah, contrary to the belief, this actually doesn't is not even connected to anything. Since we shrunk our robot from our V10 version, we noticed that there was a lot of tipping so we actually just put this under here to counteract any tipping that would occur.

[00:07:42]>> Okay.

[00:07:42]>> I see. That that that makes a lot of sense. And is that tipping like you're noticing it during auto, during teleop?

[00:07:47]What are the most When are you most noticing it?

[00:07:49]>> Um mainly it's during teleop with our long sprint or with our fast sprints to the gate when we try to rapidly stop, [music] our robot begins to tilt, which is why we also added this uh weighted belly pan over here just to combat that.

[00:07:59]>> Awesome. Let's jump into the intake itself now, moving away from that diverter. The intake as a whole is just very, very smooth. I know you guys have iterated on on it a ton this season.

[00:08:07]What do you think was like the biggest improvement you made to your intake, and how did you do that?

[00:08:11]>> Honestly, I would say the biggest improvement to our intake was or well, there's two. First, we merged our intake and [music] our transfer together using these two motors. We mounted them on one axle and had our um our axles the belted or geared together.

[00:08:25]And then, the biggest improvement I'd say is the addition of this movable part. Essentially, what it does, as we shrunk our robot, we were losing a third ball a lot cuz the third ball would come around right here and would roll off the floor. So, this uh >> [music] >> platform over here, it would fold up and squish the balls against the intake.

[00:08:45]And it also uh double it it's a double used as a climb as well.

[00:08:49]>> Yeah, let's let's see that. If you guys can just tilt that up and talk to me about how much space it saves you and what what the benefits of using it as your uh as your climb mechanism [music] are.

[00:08:57]>> Yeah. So, basically, this is it's like a multifunctional. It's powered on two servos on both sides. And so, [music] it would allow us to be able to hold that third ball very easily, which is an issue which is the largest issue we had with the intake on this robot. And so, we were able to hit that double whammy by adding it for a tilt. And so, it actually we can also rotate when it goes up, and so we can make the micro adjustments while we're up and also give us that big advantage uh in our tilting mechanism. This is essentially like we added this only for the world championship because we need we know we wanted to get those extra little points in RP. And so, we added this actually very, very late. Um but thankfully, we were able to get out the design very easily, and it worked incredibly well for us, and we didn't have any issues with that at all.

[00:09:35]>> Yeah, let's let's jump into the shooter now. You have just been so so fast the entire season. Uh one thing I noticed specifically with your flywheel is it's very very heavy. Walk me through the construction of it. What are you using for the weights and what are you using for the uh for the like a friction material on the outside?

[00:09:52]>> Yeah, so for our flywheel construction, we use 4 lb of weighted flywheels.

[00:09:56]They're pocketed on the inside cuz after worlds we saw that we needed more inertia so that our feeding time would be lower. And for our grip tape, we have cat tongue on the back and then this grip tape that we bought off Amazon.

[00:10:08]It's was lasting through the wheel or it would last often through the competition and it allowed for backspin to allow the balls to fall faster through the goal.

[00:10:16]Which we found we needed to do to beat the balls when we use the >> So, yeah, and one thing I noticed is you guys have a solid metal like it's a metal core the entire way. You're not using like a hawk back in the center, rhino wheel in the center. Did you guys Have you guys run that the whole season or did you move away from those uh you know more traditional flywheels?

[00:10:33]>> No, [music] at the start of the season we were running originally a GoBuilda rhino wheel, but then we found that this would not only give us a lot more inertia, but it allowed us to make our custom design a lot easier with just a 3D print connecting all four metal plates.

[00:10:44]>> Yeah, now talking about shooting around the field, your hood is fixed and that's been something that's been like very surprising to people all season, but you're still so consistent. How are you able to maintain that shot consistency with a fixed hood angle um in this game?

[00:10:58]>> Yeah, honestly we just control we control mostly just our power rampage on this and we use a bang-bang controller.

[00:11:07]>> Yeah.

[00:11:07]Oh, okay.

[00:11:08]>> But we use bang-bang controller as opposed to PID cuz we found it gets us to our points faster. So, it allows us to be a lot faster and it keeps our consistency high while shooting.

[00:11:17]>> Yeah, with with that bang-bang controller, how fast are your loops typically running in teleop? Like are these like 60 hertz loops, 100 hertz?

[00:11:23]Like how how fast are you able to run all of this?

[00:11:26]>> I think it's about 100.

[00:11:28]>> Okay.

[00:11:28]>> I'm not really too sure cuz I'm I'm not programmer, but I think it's 100.

[00:11:31]>> Okay, that's fine. Now, talking more about driving as well. You guys are just so so fast on the field. What automations or simplifications or enhancements have you made to your driver protocols in teleop specifically?

[00:11:43]>> So, what we did was we want to make sure and we learned this from like previous seasons as well. So, what we automated was we have this pinpoint where automatically has a position of the odometry pods and we're automatically able to track the goal by using the odometry pods and we don't use any limelight or vision at all. Which is really useful for the drivers as all they have to do is press few buttons to go to the gate and automatically just go into the shooting position. For example, we have [music] just few buttons that controls multiple parts of the system. For example, right trigger can control the intake and diverter so that whenever we go to the gate, all the driver has to do is aim it to the gate, intake, go out and shoot, which is around like three buttons max.

[00:12:22]>> Is your is your driving to the gate automated or is that manually done by the driver?

[00:12:26]>> That's manually done by the driver. What we tried to do was to make it automated but due to some inefficiencies and the lack of consistency, that's why we decided to remove it. And adding on to consistency, what I think about us as a good [music] programming team is because we like to have it simple and consistent. For example, we don't overdo with sensors and I think the only sensor we have is the odometry odometry pods itself because it gives for consistency for example for our auton. So, we had for example a limelight and a voltage drop sensor which detects the the motor voltage of whenever it takes three balls in auton. But due to it being less consistent like like less than 90%, we dropped the idea because we want the auton to be consistent. As you can see, we are right now ranked number one in auton OPR, which is why we're really simple when it comes to programming and really consistent every time as we do so many trial and errors.

[00:13:15]>> Yeah, last thing I want to ask about is auto adaptability. You guys are running a lot of different auto programs. They seem to really always work well with the partner. Can you walk me through how you're able to do that and what things you have built into your driver station for that.

[00:13:28]>> Yes. So, we start off with our main driver. Uh so, we have this configuration over here, which our auton is uh basically configurable. So, we have a lot of different types of far partner autons, where they want to have a different path. For example, if they want to only cycle back and forth from human player station. So, we make sure we have a configurable auton where we can do 27 piece from close, 24 piece from close, or 21 piece from close, which is easily adaptable for any type of far team possible. As we are uh like a close bot. So, >> For pathing, what are you guys using? Is this Pedro pathing, Road Runner, something custom? What do we have going on here?

[00:14:01]>> We are using Pedro pathing. Uh so, what we use is a Pedro pathing visualizer, where we map [music] out our location and points. From there, we use Android Studio to make the Pedro pathing uh autonomous uh thing. And then with that, we have the configurable auton, where the driver can press any buttons during initialization, so that they can make sure that we are uh compatible for any far robot, no matter what. Which is why we have a very consistent auton.

[00:14:23]>> Awesome. Uh it comes Solar. Thank you guys so much. This has just been such a fantastic interview. Every single season, you build so, so fast robots, and I'm really glad we get to take a deep dive into it on Behind the Bot.

[00:14:33]Reporting for Fun Robotics Network, I'm Amol Bhawsar, and this is team 19564 at Kan Solar. Thank you.

[00:14:38]>> Thank you.

[00:14:41]>> Thanks for watching. Don't forget to like, subscribe, and click the bell to stay up-to-date on future fun videos.

[00:14:46]FRC Tees managed store program is an easy and profitable way to fundraise for your team or off-season event. FRC Tees will handle everything from production, packaging, bulk or drop shipping, payment collection, and accounting.

[00:14:58]Teams can set their own prices, and FRC Tees will take care of the rest. Fill out a quote request, or get more information when you schedule a callback at frctees.com.

[00:15:07]Take on the decode season with AndyMark.

[00:15:09]FTC teams can discover great components such as AndyMark's 3-in mecanum wheels, [music] programmable servos, sensors that detect distance, color, orientation, and many more solutions for your teams. Find this and more at andymark.com, and count on Anymark for the reliable service that teams expect.

[00:15:25]>> To destroy