We never have to turn 180 degrees to face the basket. We can just flip that, and then our scoring system will be on the back side of the robot to prepare us to deposit it into the basket. Other than that, we also have some hooks on that front side of the robot, and that's to get us to hang on those rungs. But we're currently still trying to design that, because to get to that highest one, you first have to start off from the level one bar. And we're trying to think of ways to first hang on to the first bar, lift ourselves up from the ground, and then proceed to get on to that second one, and then pull ourselves up. We're still thinking of ideas. But for now, that's what we have so far. We're just going to be trying to get on to that first bar, and that's how we're going to do it. And those are powered by strings that will reel us on to the bar. Other than that, we have a mecanum drive chain. I don't know if you guys know what mecanum wheels are, but it's kind of, OK, but basically, it's like a wheel that allows us to kind of straight this way, but also drives forward and backward, and just makes it easier for us to drive around the field. And yeah, that's our current design. If you guys have any questions or any comments about what we've done so far, it's pretty early into the season, so a lot of stuff, we're still in the process of building this with the parts, like with real parts. So we're still kind of half figuring stuff out, but if you guys have any questions, feel free to ask. How does a wheel look like? The wheel on the mecanum wheel? Yeah. Do you want to show it? Wait, can you grab it in the front of you? I'm going. So it's like a special type of wheel. I don't think there's any use for it outside of these robotics environments, because there's really, like, it only works when you have a special type of ground. So this is kind of the thing we're using, where it's a wheel, but it also has these, like, yeah, these are, are they like casters? It's just like rollers that allow it so when it turns, it goes forward. But when you program it into a combination of different movements, then you can look. Yeah, when each wheel turns in a different direction, you can get it to strafe laterally, and that allows us to do a lot of stuff, as well as you can also make it pivot and stuff like that. Yeah, so that's what the wheel does. Does it look like you have a cost limit or a weight limit? There's no weight limit. There's no weight limit. There's length and weight limit. Yeah, Kelly, FTC robots have to be built out of kit parts. It's designed to keep the cost. Like, it's designed to keep things under control. You mean keep the kits? Yeah, there's a list of parts you are allowed to use in FTC. It's like a more constrained version of FRC, because the FRC robots are utterly insane. OK, OK, OK. I could not win an FRC competition, and most MIT engineers can't. This is interesting. Is there a bonus for the autonomous stage? No, I don't think this is. Yeah, so if you score a sample, the points are essentially doubled. I see. So it counts towards autonomous and autonomous. Remind me what the balance is. It's 30 points if you can hook yourself onto the second ladder bar. How many points is it per sample? It's like eight on the highest basket. No, with 10. Eight on the highest basket, and 10 for hooking onto the second ladder bar. So if you can score, it's 20 points per sample if you can score a sample during the autonomous stage. What kind of computer are you allowed to use on the robot? We only just have a control lab as the computer on our robot. It's just pre-programmed program with sensors like cameras. OK, so you're not allowed to bring your own compute for the robot? No. Lame. It's pretty restricted. Lame. No machine. Not a lot of electronics. Lame. No machine learning. Mostly for autonomous, we're mostly relying on cameras right now. Yeah, yeah, but what are you processing the camera input with? We're processing the camera with the control hub. All right, what's in the control hub? It's like a computer. Do you know what the computer is? Does it have like tensor cores? Does it have several cores? One core? The ability to run neural networks? Let me see FTC control hub. Well, I think we haven't really looked into that. We just assume that whatever information we give it, it'll probably get spit out. Let me go investigate. Yeah, so basically for autonomous, we use a coordinate system. OK, so it's called the Rev Control Hub with a built-in. OK, OK, the control hub runs Android. It's affordable. It's affordable. Rev Control Hub specs. Aha, all right, all right, we're making progress. The Rev Robotics Control Hub contains the RockChip RK3328. Making progress. Off we go to the data sheet. How many tensor cores do we get? OK, RK3328 is designed for a 4K television box. Support 4K, 10 bits, 30 FPS. Dang, 4K, OK, OK. Quad A35, 450 MP2, but no MPU? No? Guys, guys, what are you doing to me? This is 2024. Get a Mali 450. OK, interesting, anyway. I see. So the current plan is to have a linkage-driven, so the two yellow hooks, you're going to attempt to go for the top bar? So basically, we're trying to, so how it works is, I didn't really explain it that well, but those hooks are attached to that gray piece via a magnet. And our strategy is that we're going to first elevate that vertical slide, and then extend outwards. Then we can reach onto that bar. We're still touching the ground, so it's not an empty hang. But then what we're going to do is we're going to detach it. So because we're using magnets, it's detachable. So we're going to retract that linkage back in, which will cause those hooks to basically detach from the robot. And then we have a spool powered by motors that will just reel in those hooks, and then hopefully we can go to the robot. Yeah, OK, let's see. I mean, the overall concept is reasonable. The way these hooks are mechanically set up right now, on these drawer slides, there's going to be some floppiness issues, I think. The thing at the bottom is the mechanism detail, right? You're basically hanging off of the single supported drawer slides, which is going to have some floppiness. So you definitely want more support on the hooks. The side towers are, are these side towers structural? Yeah, it's connected onto a U-channel, if you can see. Yeah, are they, these are acrylic? It's plexiglass. Yeah, OK, so that's acrylic. That feels a little bit floppy. We haven't, we started building it in person, and so far we haven't had too many problems concerning with the slide. It seems OK. Yeah, my concern is that once you start putting some load on it, because the whole robot hangs off of the slides eventually, I'm worried that you're either going to have some, like the flexion is going to lead to some mechanism or the other jamming. It might be OK. Maybe I'm overthinking it from a wrong angle. The closest thing I've done to this is BattleBots, where everything is built very dense. So like the general floofiness seems reasonable. Do you get 30 points every time you pull yourself up on the hook? No, you can only hang once. So it's 10 points per ball, but only 30 to lift yourself up. So aren't you better off trying to put balls in baskets? Yeah, we're still trying to figure out if it's even like economic to attempt. Yeah, I don't know how thoroughly FTC thinks through its scoring rules. It's often better to, like in the robotics competitions I've seen, it's sometimes better to focus on one highly optimized strategy. But FTC is like a serious competition, so they may have a mechanism to deal with this. There are no mid-season rule changes, right? Like the rules don't allow for mid-season rule changes to fuck you over. Have they done it before? Not in the Olympics? This is like a thing that they sometimes do in lesser competitions, where if one team looks particularly successful, like it's deleting the competition, they will change the rules to ban that team. But I don't think FTC does it because the stakes are too high. Like basically millions of dollars are at stake because people's college admissions are at stake, and people spend like hundreds of thousands of dollars to try to get into college. So I don't think that's a huge issue. Because I feel like it seems absurd to build a mechanism that can lift the whole robot when you could instead build a mechanism which picks up the pieces and puts them in a basket. And the technology to like pick up pieces is a well-solved one. Grippers are a thing in 2024. So if you go do some careful research, how long do you have to build this thing? I think competitions start in November. Yeah, around the end of the year. That's right. FTC bones you by not giving you time to actually build the robot. It's actually not bad because we started in September, yeah. Yeah, oh, that's rough. And it's like part like. So if you just put like a robot arm that can just like go through the second level or something. Yeah, the arm itself is pretty straightforward. What materials, the only raw material you get is acrylic. Well, here's the thing. The previous seasons, we went for aluminum sheets and then CNC those for side plates. But we ran into some problems where our robot became really heavy. And then like hanging was a lot harder because of that. So this year, we tried to go with like lighter materials. And maybe that cost us a little bit in like the flimsiest, flimsiest of our design. So we're still trying to look at that. Oh, yeah, there's our robot, by the way.