The segway is basically one of the more advanced problems of control theory, the inverse pendulum. Think of balancing a pencil on your finger. Even IRL it's not really trivial to get it right. Seeing what you managed to achieve with the tools SM gives you is impressive.

@BeefIngot

8 ай бұрын

It's hardly complex at all. It's a basic pid loop. Get to multiple link inverted pendulums or ones with orthogonal degrees of movement

@Psycorde

8 ай бұрын

@@BeefIngot Yeah, what he said

@olivereisenberger7215

8 ай бұрын

​@@BeefIngotit's simple in theory, but getting it to be stable and controllable is difficult.

@BeefIngot

8 ай бұрын

@@olivereisenberger7215Nah. Like I was joking somewhat in the smugness of that reply (didn't come through I guess), but it actually is pretty simple, and there are all sorts of auto tuning algorithms out there right now. Even still, believe me, because I've implemented them myself in lua within stormworks, in real life with a microcontroller and 2 common dc motors and more, that it really isn't all that complex. When you look at the wiki, if you haven't been in math class for a while it looks like gibberish, but if you go slow and read what each of the terms does, its actually fairly easy to recreate that formula in your programing language of choice. Heck, you can even do much more fancy things than the base formula if you just get a basic understanding.

@isntthatnice.2148

8 ай бұрын

​@@olivereisenberger7215thats what pid is for

0:16 i love how scrapman still continues while that first segway just slowly rocks back and forth and getting more aggressive at it with each rocking 😂😂😂

@kj25

8 ай бұрын

It even does a flip here 2:16 😂😂

You should do stuff like this with Kan, it would be fun to see y’all each build your own versions.

@BrentStarling

8 ай бұрын

I think he and kAn just have different recording schedules, so they don’t do much multiplayer outside of MM.

@rojnx9

8 ай бұрын

kAN would try to use two piston engines to power this, which would actually solve the variable engine power problem.

You should try making this in trailmakers, because it is easier to program and the wheels are bigger in comparison to the blocks and parts, and logic gates can output different amounts of power to the engines.

To get varying speeds, you can put another engine on each wheel that also goes forward and is connected to a sensor that detects the ground later. When you move the stick a little bit, one engine is active, and when you move the stick more, both engines are active, making it go faster.

@JF743

8 ай бұрын

I was going to say that. Analog movement is extremely important for a segway, ideally you'd feed gyro values in the motor, but considering the limitations of SM, having 2-3 set speeds would help a great deal.

@loganroman5306

8 ай бұрын

The mod pack logic has some tools that might be ideal for this project, assuming you were ok with going modded.

@foamyesque

8 ай бұрын

Only problem is, then you're looking at four bearings per wheel, or more. Might need to do piston pass-through shenanigans to keep the axle rigid.

You should give this challenge to kan and see what he can do with the smart engines and number logic.

more scrap mechanic, finally.

@turtleo4732

8 ай бұрын

He played scrap mechanic yesterday

@video_factory_UA_offical

8 ай бұрын

​​@@turtleo4732 WHEN

@turtleo4732

8 ай бұрын

@video_factory_UA_offical what part of yesterday do u not understand

@wingid

8 ай бұрын

It's been 10 days of no scrap mechanic besides the hide and seek

@turtleo4732

8 ай бұрын

@@wingid so scrap mechanic

I love how the first Segway just slowly flipped itself over as he was demonstrating

My only thought for variable speed is extra engines and a bit of logic to select how many are running at a time on each side. Gonna get bulkier quickly though. But that could help with both easing into a direction of travel, and with turning if programmed to slow one side.

I think you should increase the wheel size and weight not only will that increase speed but also gyroscopic percussion that will increase the overall affect the wheels have on the creation

I'm pretty sure the reason the reverse is so bad is because the handle is positioned toward the front, so it needs to tilt farther back than forward to create the balance shift to tilt it backward.

@jurare89

8 ай бұрын

Yeah. Also: the center of mass on the creation is entirely different than in reality, since we dont have a person standing in it, but just the handle changing position forward/backward. While a real person would not only lean forward or backward, as well as left and right, you would also lower/raise your center of mass based on acceleration/deceleration.

@Vessekx

8 ай бұрын

@@jurare89, I wonder how much it would help to build an entirely passive ‘person’ (hinged at ‘ankle’, ‘shoulder’, & ‘wrists’) to get the center of mass from the weight shifting a bit more centralized.

@jurare89

8 ай бұрын

@@Vessekx I´d assume it massively changes things. But while it would probably help massively with overcompensating/undercompensating, it would also be insanely complicated to program. I don´t think it would be impossible, but it would take a gigantic amount of programming, experimenting and adjusting to pull off

I think it could work if you had more motors set at different speeds for each direction, but it would be huge and hard to wire.

@lunlunnnnn

8 ай бұрын

An easier way would be using modpack PID controllers, which is actually the same algorithm as what IRL segways use to balance

@tylercollier9703

8 ай бұрын

Second set of sensors that trigger when the creation starts to lean to far.

Couple of ways you could mitigate some of that oscillation, maybe. Make a larger build, and add a second set of motors at higher power, and a second sensor set to enable them when the thing leans farther than some angle. Second option is to add a second sensor set on a short timer with AND gates, and two more motors. I actually don't remember if you can hook multiple motors to a single bearing, which is what both ideas kinda hinge on (unless you add more bearings). You might also have an extra set of bearings on controllers that adjust some rate based on multiple sensor inputs.

Scrapman makes everything in both games even segwayes

I love how when i started watching you're channel, logic was something you struggled with, but now you're really starting to get the hang of it :)

2:33 that one Segway just vibing in the background xD

Thank you for including my Segway in the video.

Segways are such stupid creations and I've never loved learning about them so much, thanks scrapteacher!!

You can take the man out the scrap, but you can’t take scrap of of the man.

You should try this again but using the PID from the Modpack. Since segways IRL also use PIDs. Could be a cool learning experience if you don't know how to use PIDs.

Now I haven't played this game, but maybe you could use more engines to essentially make it ramp up the speed. I'm thinking multiple sensors in both front and back that activate depending on how much it leans forwards/backwards. When it leans further it activates the faster engines (not sure if it is a problem when you also have the slower engines, maybe they need to deactivate at the same time). The build would probably be pretty big though.

Scrap mans segway flopping back and forth at the start in the background is hilarious

I thought this video was going to be a segway pun into another idea but this was actually really interesting and cool

We love you scrap

2:17 the toilet segway did a back roll 😂😂

the biggest problem with segways irl, is how easy they break you ankles. Simply wanting to step off the thing will cause a shift in your centre of gravity, meaning the thing will want to go exactly where you are about to place your foot

Honestly might be more doable for the wheels with a quad-baring set up, that way you can have two sets of sensors, one for slow moving, and then a "backup" set for when it starts to tilt waaay too far

It works the exact same way as a hoverboard. You lean forward to move forward. You lean backward to move backwards you lean one foot forward and one foot backwards to tank steer

Multi-player Monday Segway races would be hilariously bad!

I have a sugestion when you are doing hide and seek maybe use the seeker view while you are hiding

You should try this in trailmakers it would be interesting to see if it would be easier or harder to make it work like real life

an idea of how to make it work, use weighted levers. almost like the levers in pin ball, you have them pointed toward the center with weights on the end and then when you press forward they swing out toward the front and then when you press backwards they will come back to the center then swing toward the back. that will help to keep the weight over the center and lower toward the base and you can change the position of the weight quicker

There is a way to do the steering and drive forward, reduce the Reverse motors by 1 to 2 ticks (the only problem in this is the reverse is slow and the reverse and turning gets messed up a bit).

the wedges really helped improve the look and overall professionalism of the segway, maybe adding a few more would make it balance better because of...... wedge science reasons

This would make it much bigger, but you could have 4 engines per wheel. 2 low speed (one forward, one reverse) and 2 high speed (one forward and one reverse). The low speed engines activate when the angle is shallow and the high speed engines activate when the segway is leaning a lot. Then you can balance the speed of each set to improve the stability.

love the vid

Scrap mechanic has the ability to do variable speed in this context, it just will be a lot more complicated and larger. You just need more bearings, more engines, and more sensors. One sensor could be set to a higher number, and will activate the slower setting motor, and one sensor would have a lower number, and would activate the faster setting motor. You could expand on this indefinitely, with more granular motors and sensors, as long as you were ok with it being really huge. You could even have multiple motors working together, so when the second sensor toggles state it would just have 2 motors with the same settings activated, thus increasing speed.

@PDsPCRepair

8 ай бұрын

Say you have 4 sensors for forward. Each of them at progressively lower ranges. Then you also have 4 motors, all at the lowest setting. When the first sensor activates, it would trigger 1 motor. Second sensor activates from leaning even more forward and it would trigger the second motor, and since the bearings would be working together, it would increase the speed. This could happen progressively and accomplish the variable speed you were saying could not be accomplished in Scrap Mechanic. This would take up a lot of space, but definitely possible.

You might also get a tighter feedback loop for the sensors if you put them on fast pistons moving out of not triggered and in when triggered. The angle to which way you're leaning should make it occilate as moving out brings it closer to the ground.

this has worked out great. the only thing that can solve the obvious issues is variable speed engines. maybe a bigger build where you can put in more details, regarding functionality. i actually never played scrap mechanic, but maybe you can let a gas engine take over as soon as you reach the desired angle.

You have 2 wheels so you could use one for balance at low tilt (staying uptight) and 2 for high tilt. Also if your turning mechanism physically moves your left and right height sensors you could effect a bias between forward and backwards seperately between the wheels.

I like what you did but what I probably would've done is set the engines connected to a controller and logic gate setup where if you are holding a button for a direction, it'll increase the engine's speed to create that direction

Finally back to Scrap mechanic (MM doesn't count)

Segway race for multiplayer Monday? Think it would be fun to see you guys try and master that thing while racing each other.

It would be trivial to get it right if we had an angular velocity sensor. You need to know the difference between a vehicle that is upright and stable, and one that is passing through being upright at 1000mph.

Might be complicated but a way you could try the slow acceleration is put different sensors on different engines that would speed up with the smaller sensors.

Greatest youtuber

You should try doing this in Trailmakers, since that game has angle sensors and you should be able to get more control over the engine speeds with the different logic system. I know that game has its own weirdness, but that's why it's always fun to see the same concept done in different games.

scrap mechanic could really use a speed controller, controlled by logic blocks to adjust electric engine speed. would greatly improve building capabilities.

For "variable" speed, you could maybe put the wheels on pistons. It will move away from Segways though... But if they're in the bilaterally symmetrical you get what you have now, if you pull one closer to the center of mass, you'll change the radius. So even though it's spinning at a fixed speed, the net effect is variable travel. This only really solves the merging of balance and steering though.

This reminds me of your World of Contraptions Series for some reason

The only thing I can think to change about this build might be the weight distribution. Even with the concrete that you added below the base, a real segway's base platform is actually multiple inches lower than the center of the wheels, so your balancing system might be working against itself slightly due to the fact that your base platform extends so high above the axle and that your engines are in-line with the axle If Scrap Mechanic has biggers wheels, maybe give those a try so you can lower the vertical weight distribution. Otherwise, you might just need a lot more concrete under the base to compensate for where the weight currently sits

Maybe use 6-4 controllers on a loop instead of 1 engine. I remember seeing some building a car using controllers in SM survival to save fuel. Having up 2-4 sensors that detect different distances plus controllers that have different rotation speeds on a loop could possibly create a variable speed mechanism. Just a quick idea that I thought you could try in a part 2 of this. Otherwise another (non-scrap) video from scrap man. Great stuff. PS: One of the best things about your content is the fact that you (some who actually cares about the community and isn’t just about the money) makes it. Recently I’ve rewatched some of your older highlights videos one of them being the scrap-pocalipse.. (as I KZread only kid at the time I remember I was really confused by it all) glad you stuck around. Your a good person and it’s just easy to see that in your videos (the way you address yourself to us as “ladies and gentlemen” and instead of saying the trendy “Let’S hiT 4,0o0 lIkEs today” you just humbly say “I hope this video has earned your subscription”) and not only did it earn that but my loyalty as well. Ever since I’ve first stumbled upon this amazing channel I’ve always considered you as my #1. Looking forward to watching tomorrow’s video.

@StephenHall-zz2ym

8 ай бұрын

Using angle sensors in Trailmakers would be a lot easier.

You could have different engine speeds, just gotta use 2 or more engines per wheel and direction. You could even use one speed as a kind of base speed and then add the second speed on top for turning or more correcting

multiple sets of engines perhaps, have one set of engines for lower speeds that will work for in-place balancing and a one or even several more sets of engines for higher speeds programming may be a nightmare to choose which engine to activate at the appropriate time though i believe it should be possible, one issue would be the size of the creation though

IRL you have the wind helping you not lean more unless you want to. That combined with the engine limits makes it harder

more engines would give you speed variation and more sensors with different trigger distances would give you the variable speed and balance controls you are after no? Appreciate it would be a much larger Segway to accommodate all that though!

I think it's just need to be bigger so more components can be used, mostly motors set do different speeds linked to more bearings, might not get gradual increases but at least 2 or 3 steps depending on how many bearings are activated.

For the balancing try have sensors with different ranges and have them input as pulses and asxit goes closer to the ground have them pulse faster and faster to have more to cominsate

you can put multiple engines and program which one is working depending on the angle, also multiple sensors ...

Its not about the speed, it's about acceleration. When You reach constant speed you will continue leaning forward because Your wheels will be before Your center of mass and gravity will do the work. I'm not 100% sure but I think that in real life segways You lean forward to accelerate and lean backward to decelerate. Of course it's more complicated because of air resistance and the center of mass is low due to heavy engines being at the bottom of segway so leaning the stick forward has lower impact and it's of course tuned by software and some gyro but the rule is still leaning forward for acceleration and leaning backward for deceleration, not for constant speed and that is why it is so difficult in SM.

Good vid

i think if you take this exact build and just make 2 more engines and make 2 more barings for the wheels and then 2 more censor in the fron and back but these 2 new censors are gonna be for the 2 new engines and barings so that way it goes even faster if it reaches a certain point and that way it corrects it self when going too quick example lean this much: / (only 2 barings are going at the same time) lean this much: - (all 4 barings are going at the same time with combined speed) i think thats a very simple fix, idk xD

Also segway hide and seek would be funny handi cap for everyone 😄

I just learned how a Segway works and it blew my mind

how about making a challenge for viewers for the best segway bulit in SM? , to make this interesting maybe you could build an obstacle course with ramps, jumps turns etc.

the only thing i can think off is trying the modpack so you can use number logic and the smart engine controller which would give you the ability to vary the speed based of angles if im correct..

Its because they dont stabilize like that when you push forward on the handle bars, and when you stop and pull it back the tires slow down, and they press on the brake a little and everytime you pull the handle bars back while moveing forwards, it automatically slows the tires down so it can gwt under your center of mass and stop you from falling backwards

What if you try to make this same idea in main assembly or roboCo, where the programming system is more complex and allows for these types of complex behaviours?

I picture a pair of pistons, one on each side of the bar with a matching set facing backward. Each has a weight on the end of them and they extend forward to smoothly lean forward and then the rear facing extend to go in reverse. Would that work?

I think this still isn't a real segway, the handlebars don't actually bend off the wheelbase, and the person leans forward and changes the weight distribution, not sure if it matters, but it could since the bars don't go back to straight as the wheels try to push themselves under the centre of mass, they stay out slightly in front. If Scrap removed the handlebars tilting forward, and just kept the front sensor angling when he pressed the forward button, he could create an actual segway and could have speed ramp control by making it loop and having the seat increase and decrease it or something. although there might need to be some kind of kickoff momentum change to get it into a balancing back-and-forth state. also if Scrap doesn't mind about how it looks, and more just that it functions like the real physics of a segway, he could add a (somewhat bulky) system with suspension and thrusters or something to dampen the system like in hovercrafts (I think a controller bending the sensor out of the way could also work, or a piston to slowly pull it back or forward when it activates, and yes real segways don't have a suspension thruster system, they have compact computers calculating how much speed to put, but its scrap mechanic and it does the same thing so I think it would be appropriate to add). Another addition that is bulky is a controller created turned circular piece in the front that has 3 separate sensors at different configs that activate different engines or propulsion systems (to have different speed abilities for different lean distances, would require double or triple bearings, maybe running a single axle or something with an attached piston passthrough used in piston engines, could also remove the double bearing tank steering system and use those bearings to incorporate an increased speed to one wheel rather than a reverse and forward wheel). (disclaimer these are vanilla solutions I've not tested or anything, just thinking off of my SM knowledge. Also could all be solved in about 5 seconds with the mod pack or something) (also would love to see a piston engine version of this, since they do have ramp speed capability, would be so bulky tho)

Gas engine-bearings are fixed when controlled by logic as a gas-engine locks when off, thus they can be used for tank-steering.

Going to have to test but could you use switches to increase and decrease forward / backwards in 15 degree increments

You could try using the Fant mod speed controller for the engines.

wouldn't you be able to get tank steering if you put the baring on the wheel with another wheel on it? I mean there would be more drag but wouldn't that work?

Just a couple hundred thousand subs till a million 👍👍

It would be quite interesting to see how your design could change functionally if you built it in Trailmakers

I guess you can do triple bearings and turn on 3rd ones in steering. 3rd set of bearings would be powered by a higher power setting engine.

To maybe fix balancing problem add a motor with a higher speed and a sensor if it detect that it is leaning too much that motor will kick in

You guys should do a build challenge where you see who can build the most durable plane

i think it can work if you add a 3rd bearing on each wheel so that insted of just stopping a wheel when you try to turn at speed it also make the other wheel spin faster.

Yes need numbers on shocks n engines

why not use electirc engines for turning and gas engines to go forward and backwards?

Put a big guy (or buster?) Of varying weights stiffness floppiness or springyness on it and maybe it will stop the wobble. Or maybe it will turn it into a self driving flailing chaos machine.

Maybe a third bearing powered by an inactive gas engine can be used as a sort of clutch to reduce the acceleration of the wheels.

quadruple stacked bearings so you can have more engines on another bearing and sensor for going super fast if you lean a lot and steering can make one go even faster

asking scrapman to build a machine that pushes one building into another one, like a domino, in instruments of destruction

Maybe its possible if its built to a massive scale to contain many engines for turning states, further leans etc.

Do a hovercraft Race around the airborne map pls❤ I watch you every day

I have an idea: What if you try to simulate the ramp up speed with pulse systems? Like have 2 sensors each side and if the more angled sensor gets triggered you give inputs via a pulsing logic and if the other sensor triggers you make the input direct. Atleast you could make some steps.. idk how or if this possible but I think its worth a thought

@Der_Yoloist

8 ай бұрын

Also another idea: one problem was the aprupt braking. if you but the wheels on sideways suspension in direction of drive you may compensate that

Can the electric motors be ran on PWM logic? It was our go-to for basic motor controls in electrical engineering classes.

I wonder if setting up a pwm system with the motors would give you better control.... It depends on how fast the motors can be switched on and off, and how much momentum the wheels would maintain/lose between the pulses... if you had a very fast clock that would be able to turn the motors on and off for different duty cycles it might give you control over the speed. (Theoretically with no losses a duty cycle of 50 percent would be something like every other tick would power the motor, and generate 1/2 speed).

Scrapman idk if you can do it but use logic to reset the weight (handles) back to folcum point to stabilize? And possibly for movement turning try using logic to have 1 wheel on temperary(using delay) for movement turning and have it unactived at stand still in the balance position also use center counter weight not long counter weight can cause wobble loop law 2 force in motion stays in motion. ( basically have a reset position or no movement have neutral wheel engagement when at speed (can use speed logic at speed logic on at no speed logic off) and reset balance on release of controls wheel go off but balance goes to center if that makes sense sorry hard to explain for me.

CLAIM YOUR FIRST DAY TICKET BEFORE IT’S TO LATE 👇

@ScrapMan maybe you could have added an extra set of engines that kick on once the vehicle leans past a certain degree to keep it from falling over?

I think something missed is that if the Segway doesn't fall when off, it's not actually the same as a Segway.

alls i can think is triple bearing so you can have added speed at certain tilt but it would need forward and back on said bearing but idk how it would differ sadly cant test it myself have to play with phis 7 or slowmo even with basic car so double bearing isnt something i can do myself its to floppy

Quad-bearings would let you have multiple speeds. Triple would let you have a unidirectional ‘boost’ for turning. You’re basicaly trying to tune a PID-loop where you don’t have access to P (period) & D (duration), and can only choose a single value for I (intensity). You can ‘fudge’ the D value a bit by tweaking your sensor tilt, but you’re limited in even that. Not sure why you couldn’t use basic steering bearing setup for the left/right, though, instead of double-bearings there.

Bigger wheels would help and mods like the smart engine can ramp up, but then it would require some programming and not be vanilla.

I'm not a pro player, I'm not worthy to give opinions, but I have an idea. I'm sorry if it's silly. With some modifications, and without making it bigger, I think there's a way. So, here's the idea Bearings needed per wheel: 4 Propulsion: Try controllers instead of engines. total of 8 controllers  - 4 for forwards   - 2 for each wheel.    - one set on loop (slow)    - one set on loop (fast)  - 4 for backwards   - 2 for each wheel.    - one set on loop (slow)    - one set on loop (fast) Sensors: 8 in total.  - 4 facing front   - 2 for each wheel.    - sensors should be   placed close to the      front of the wheel.       - one looking down       at a low angle       - one looking down       at a lower angle  - 4 on back   - 2 for each wheel.    - sensors should be   placed close to the      back of the wheel.       - one looking down       at a low angle       - one looking down       at a lower angle Additional counter balance: Try Pistons (placed one a block which is tilted at a small elevated angle from the ground-level) with weights to counter below the system by 1 or two blocks. And some logic blocks: 8 in total:   - 4 𝑂𝑅 gates   - 2 for forwards    - 1 for each wheel   - 2 for backwards    - 1 for each wheel   - 4 𝐴𝑁𝐷 gates   - 2 for forwards    - 1 for each wheel   - 2 for backwards    - 1 for each wheel I'll continue to write ✎ if I get time, I've an exam tomorrow forgive me. And please wish me luck

One idea to at least try to get around the engine speed problem would be smart engines and number logic, but in vanilla you could do 4 engine speeds with quad bearings and 4 motors per wheel. You could have slow only, fast-slow, fast only, and fast-slow, and then the reverse for backwards movement. This would require a ton of logic and 8 total sensors, but it would also let you move while tuning at variable radii by stepping the speed of each wheel up or down.