The thing I like about Scrapman is that he always gets there. Like, I'll be sitting here going "slow down the pistons" or "maybe take out the negative?" and he always tries it before he's done. Some builders don't show their process, but he does, including his struggles, and I respect that. Also, at the end of the day, he really knows how to build in Trailmakers. Looking forward to him honing his Screwdrivers skills!
Instead of using pistons on all four sides you can try to make a piston triangle instead of a rectangle. Rectangle having 4 sides can have a side-to-side diagonal wobble issue, but a triangle won't have that issue, it is like how tripods for cameras have three legs only, this prevents the wobble issue. you can use anchor blocks to attach to the ground to fix the drifting and to prevent the vehicle falling due to instability in a tri legged leveling system
The problem with four pistons is because it doesn't know whether a piston is in contact with the ground. So the negative inputs try to retract a leg to lower that corner, but because it's not on the ground yet, it's lifting it up when it should still be going down to make contact.
@@TheBaldrickkYes, I think most of his problem could have been dealt with by using a manual input to fully extend all pistons and then set the automatic system to retract some of them untill you are level. That way you would eliminate all the random concacts with ground messing everything up (like the wheels touching the ground sideways while trying to level side to side because the pistons are still too short)
@@qdaniele97 not really. A tripod will always have all legs in contact. You adjust the front two for side to side, and then the rear for front to back.
Like all things with legs, if set up properly it won't wobble. Doesn't matter how many legs. The argument that three legs is better than four is absolutely absurd
The simple solution was just to have a tripod system. Two in the front to manage left/right and one in the back to manage front/back. Any three points will always be coplanar, so there won't be any wobble.
Multiplayer idea: Teacup race. Each racer has to make a vehicle that will keep a loose piece of cargo on top of it as you race around the desert map. Losing the cargo means you have to stop and respawn the cargo (aka the teacup) before you can continue. Extreme acceleration in such a race has the obvious drawback of making it harder to keep your cargo on your platform.
Good idea, but they need to use a standardized teacup, or else the winning strategy is just 90 degree walls, more like a coffee mug. meaning it never spills unless you almost tip over. Rather than respawn he cargo though, I think a crane attachment that needs to retrieve it would be better.
This might be a cool idea Post 1.9 Update with the new Vehicle Collision Dynamics, then the Cargo can't just rip the vehicle apart from the inside of its Cargo Space from just merely innocently rattling around a bit.
19:05ish... ScrapMan: "I wish we had an Emergency Brake." Incoming 1.9 Update: "Wish granted!" 😘 26:55ish... Scrap, I just got an idea! You spin your entire Missile to spin just the Dynamite... but spinning the Missile throws it off a bit. The Missile never needed the Spin, only the Dynamite Delivery does. Therefore, I propose an idea: Separate the Body of the Missile into 3 Segments with the separation points at just above and below the Missile Body Parts holding the Dynamite. Place 2 Rotating Servos in the Body of the Missile, one on the top and one on the bottom of the Dynamite holding Missile Body, and set them to spin the Dynamite to dispense it while the Top and Bottom Missile Body segments do NOT Spin. Crazy or Crazy like a Fox? :D (Also, wasn't there some kinda Hack where some Part or other was placed beneath the Rotating Servo to trick the Torque into getting Walled there and Gandalf-style "YOU SHALL NOT PASS!!!"...?) End of Vid: I wonder if any of the new 1.9 Logic will help improve this even further? Also, I wonder if Yzuei and the others should be scared for future Challenges in how outgunned they're getting in the Logic Force Powers Arms Race...? Yzuei was having a hard enough time trying to face ya in this Challenge with your Truck that did NOT have self leveling capabilities... :P
Simple Solution: Base of the Aiming Mechanism ist build on loose joints. Two 2x1 Joints facing to front, on top two facing to the sides. Now on top Gyrostabalizers set to max. And on top the Launchpad
Might be easier but it's not a very elegant solution, I'd even say this solution is borderline cheating. Personally I like to build realistic vehicles and mechanisms so I usually stay away from gyros. Good luck finding a real army vehicle that is using your mechanism! Of course gyros do exist in real life but they're like a million times less powerful. People that are fixing all their problems with gyros are just lazy with zero imagination.
@@Alfred-Neuman i dont say, that i didn't like the complicate solution. But i know your point and appriciate it. Also I have to add, not always using a Gyro should be considered cheating as sometimes a "simple" solution is the beginning of getting creative. Sometimes it's better to have a positiv but simple outcome as a negative complex, especially when you are a beginner OR need a q&d solution.
Actual IRL ICBM trucks have a deployable launch pad that folds out the back of the truck and makes the truck immobile while stabilizing the missile and they work in a combination of how scrap man did it and how you described.
I was searching for a comment like this, i was feeling so bad watching that piston activate all the time and him trying to figure out why. I was like "Its the pistoon, its the pistooon"
The uneven ground is causing issues I think, because It seems to work when the hill is flat (but not level). Try making it balance on 3 points rather than 4. A stool with 3 legs can never wobble, but one with 4 legs always will if the floor is not flat.
i heard there is a proof for non-wobble 4-legged tables on a ground with some limiteto the maximum slope the ground can have. essentially if you rotate the table at the center on that ground, there is an orientation where the table doesn't wobble. though 3 points are obviously easier, but i thought i mention it
@@Akronsus Even then, only 3 of the legs are supporting the weight, the other one is in contact with it, yes, but there's barely any weight on it. It's also what makes rubber/fabric feets on chairs effective, they can squish until the fourth leg is in contact, it's still rocking within the compression of the feets, as it's physically and mathematically impossible to have more than 3 "true" contact points, but so little that you can't feel it.
i also did a self stabilizing missile launcher and i made by using these steps 1) the pistons all extend until a sensor doesn't detect the ground anymore 2) the gyros stabilize the vehicle by lowering the pistons that need correction 3) fire the distance sensor again but only for a 0.5 sec pulse to just give it some stabilizing buffering height
Well done. I see some comments saying just the launch mechanism but in my opinion using stabilizers to level the whole truck was much cooler. Would like to see you perfect it maybe and use it on different kinds of creations.
you can make it better all you got to do is step 1: Slap a "sensor" on at EACH "leg" (da pistons dat to da stabilizing) pointing DOWN with less than 1 block DISTANCE (or until they detect the ground at the same time the "leg" touches the ground) step 2: when the "stabilizing" part is DONE use the "sensors" to CHECK if it DOES NOT detect the ground if so EXTEND the piston until it DOES step 3: now do the opposite CHECK if it DOES detect the ground if so CONTRACT the piston until it DOES NOT step 4: do step 2 to 3 again and again until its "balanced" Note that Im a programmer and been programming for 4+ years so this may work not sure since I cant play Trail Makers (yes... bad PC)
It would be difficult to add with logic just because a loop is hard to make in trailmakers. Needs both several blocks for each step, but it needs a constant input to go around, not sure if it's worth the trouble in trailmakers
I think it would be really fun to see scrapman try a coding game like "the farmer was replaced", especially because he knows all of this logic and that is the base for a lot of coding! would be very interesting to see his thought processes
I think the issue with the negative input is that they affect both pistons on the opposite side regardless of whether they're contributing or not. If you instead had sensors on the feet of the piston that went into an AND gate you could make sure the pistons never retract when they're not in contact anymore
So the biggest issue I think your facing is having to calculate the extensions of 4 points, if you were to change it to use 3 points instead (back left, back right, front center) you can designate each point to control their specific rotational axis. Front would be responsible for your forward tilt. This will also fix the issue you faced where you are resting on to points and are rocking between them
i dont think thats the issue. the main issue im seeing here is lack of data. its like giving a robot only the altitude off the ground den trying to make it stand up just base of that. the granularity of the information he was working with was jsut way to small for what he wanted to do. shoudl given each piston its own set of sensors.
This was awesome, Scrapman! The only thought I had was if you were willing to be more manual, you could take advantage of the fact that it doesn't necessarily autoretract when you kill input to separate left-right and front-back auto adjusters two different inputs. Activate the get the tires up pistons, then forward back and then left right. You would need a different cancel button, but you're gonna repair back to neutral anyway so it may not get used much anyway
In control theory, you created whats called a "bang-bang" feedback controller, where the two options are on/off for control, therefore the name when controlling something loud irl "bang on, bang off". What you need ideally is a proportional controller, where the control output is proportional to the amount off center. This is the most basic of a "PID controller", which is the standard feedback control loop in engineering
I absolutely love games with logic (like scrap mechanic or trail makers), because the only time there’s an issue in the logic, isn’t from another factor, there’s a problem that can actually be solved, wind or the ground can’t affect logic, which I just love. It’s very similar to code scripting, like Python, which also runs smoothly if there are no mistakes, similar to logic in games, both scripts and logic take everything literal, with can make them like a big, confusing puzzle.
I feel like the easier way to implement this would be using the pistons as stands to keep the wheels from wobbling and moving the vehicle, and have the auto-correction on the missile launching mechanism. Especially the front to back, instead of trying to adjust the whole car 10 degrees, why not just adjust the launch angle 10 degrees.
Idea: You guys build each other a plane to use. Example: Scrapman makes one for incognito, Incognito makes one for Yzuei, and Yzuei makes one for Scrapman? 👀
Video idea: Cluster missiles vs a very fast dune buggy thing Or against a drone/jet ( to make this more balanced I think that you shouldn’t be allowed to Actualy crash into the target, but have to launch your cluster explosives at the target) (Maybe using a sensor with a large distance that releases the dynamite at the front of the missile would work.)
I also think in this case the biggest issue is, the pistons are retracting, because once you are considered level, the pistons contract because they feel like they are no longer needed due to the angle sensor.
How could you not understand XOR? So simple... If there is no signal going in, there won't be a output If 1 signal goes in, there will be a 1 output If 2 or more go in, there won't be a output anymore
The missile's trajectory would be skewed by Magnus Effect, although I'm not sure if it applies in Trailmakers. Depending on how aero is calculated, that might be a contributing factor. Basically, a wing (via the complex shape) causes air to pass by each side (top/bottom) differently, creating lift. Magnus Effect refers to how a spinning object can create the same effect without the complex shape. If an object is rotating around an axis not parallel to the windspeed direction, one side will be rotating into the wind, while the other rotates with the wind. These two sides have the same effect as the two sides of a standard wing. The missile, while flying downward, still has some forward (sideways?) momentum. When it spins, although mainly travelling down (along the same axis as rotation), the sideways drift works with the rotation to create a slight Magnus Force. Again, I don't know if this is really a factor in the game, but theoretically, you know...?
13:41 You could let the middle of your vehicle pivot (on the vertical plane) from the middle, kinda like the opposite plane of travel from the oscillating loader, but you will need peramiters to keep it from twisting past certain points. Then, just level front and back seperatly.
So, in regards to the missile skewing with rotational force, (I never realized that was the problem till you pointed it out) I would make vaguely biologically functioning submersibles, and this being before we ever actually had gyros, I made use of helicopter servos. I would run into a similar problem of rolling in strange directions with an odd number of helicopter servos, so you just have to do it like a dual blade copter, two servos in opposite directions. If you build it properly, the fuselage may not spin at all, but you could still spin the dynamite itself around the fuselage, and then detach after timing, and have the same effect.
So just a bit of info on control theory and physics that you may be able to apply. Generally the less control operations the better. Meaning less things happening simultaneously the better. I would personally attempt to activate a single piston location at a time. Ie front left, front right, back left, back right. And allow it to balance in all directions at the same time. Just loop through that process indefinitely. I would also only give it a fixed amount of time on each piston, 0.1 seconds and then move on. This will act as a relaxation parameter and will allow the piston to move in the direction of improvement rather than all the way which should eliminate some of the oscillations. The final thing is you are correcting four, which brings nonlinear solutions into the realm of possibilities. If you figure out a tripod configuration you may have better luck stabilizing it as well.
I've dealt with this in ScrapMechanic. The way it can work easily is by using 4 angle sensors and rotating them 45 degrees so that they measure the diagonal tilt instead of coronal/saggital tilt.
scrap i got an idea. pirate ship battles 2 or 3 small ships vs over powered pirate ship. weapons cannons only and a block limit of 100 to 150 for the little ships
For a flat plane you want 3 points of contact, not 4. With 4 points you need each one to be perfect of you wobble. And programming it will make them fight each other. With three points of contact(2 in the back and 1 in the front) the two at the back can work on levelling left/right with the front pivoting to conform, then the front one can adjust the front/back level with the back ones pivoting to conform
Have two sensors per side and they only deal with that side of the vehicle. Two pitch on the right side, two on the left, two roll on the front and two on the back. That should solve some problems
One potential solution would be to have a set of pistons for roll and a second set for pitch. This would prevent the two axis from sending conflicting inputs to the same piston.
Suggestion: *[TLDR available]* Use pistons to clip in X number of Dynamite, and have each of them connected to a hinge and that connected to a coupler (disconnecter) so the hinge folds the dynamites upwards, and when you need to release it, you first fling the dynamite with the hinge going down (Adds momentum) and on a short delay, seperate them, this will make a good non-spin and high accuracy missile... Might be a hassle to make but maybe you can also do a TrailMakers MM challenge, like someone makes a missile and someone makes an armored bunker, and you can have Yzuei make the missile and you make the bunker? And the challenge is that the team that takes the least shots to destroy the bunker (destruction can mean anything: Maybe putting a buncha dynamite in the center that is covered up by the bunker and if that explodes, then you're out???) TLDR: Instead of spinning, flick the dynamite outwards with a hinge...
One issue was that with the negatives inputs included, whenever one leg is extending, the opposite corner leg will always contract (no matter if its front to back or side to side), creating the wobble. A three leg system would have been easier to manage because it would never tilt and the position of each leg is directly corresponding to the motion of the truck.
To get the wheels off the ground, an easy solution would be to just add some extra pistons which are active no matter what hwen the stabilization is on
I'm going to assume to you saw my comment on that video for this idea! :P It would've been easier to use the pistons to auto-correct the left and right angle only and adjust the front and back angle through the missile angle directly. To have it automatic both ways with pistons, the easy way would be to have a set of pistons for front and back angle only, and use the other set of piston for left and right angle only. Your system is pretty neat though, and it should have more range of motion for extreme angles. Good job!
(in middle of the video) I thought of Each of the legs having separate control diagonally X not left right and front back 🔳 so they don't fight together but cooperate 🤔 (endedit Basically the sensor needs to be not left right but ↙️↗️↖️↘️ It's should be technically the same thing with logic just 45 degrees And each need 1 piston to lift wheels of ground separately. 27:20 Why do you never do the same logic but with slower and only 0.3 degree range (instead of 2) that way you won't overcompensate which is often caused by overshooting the range of degrees thanks to momentum or and delay of deactivation Enjoyed the video hope you read my tips 👍
I think the solution would be for one corner of the pistons having 4 angle sensors, so every corner will having own x and y axis. 2 sensors will extending the pistons and the 2 sensors will retracting the pistons (for example the back left corner will having for X axis the one sensor that is 90 direction and 179 width will retract the pistons and the one with -90 will extend the pistons)
Something that could help would be to rotate where the 4 piston sets are so that there's one pison on the front, one on the right, one for the back, and one on the left. That way there's no doubleing in the programming. Each angle sensor will only output to a single piston. Thinking about this now, though if the vehicle is tilted forward and only a front center piston extends, the vehicle will only be supported by 2 pistons, becoming unstable...
Idea to make things easy: Always make a simulation of what are you trying to do mathematically in some programming language first, so you can correct the errors in there, and also, use the new gates
I had the exact problem with my missle launcher but i went with a Missle Bay that rotates around the longitudinal every Missle Pod lat angle alinged according to the sensors. The advantage is that you can fire while on the move
Hello, I've made a Harrier at trailmakers but I'm embarrassed to upload a video on TH-cam so I posted it in the gallery (the design is called "Harrier G3 (no landing gear)" it's a nice harrier :)
try to make many different launch programs such as 100m 200m 300m 400m 500m so that you just push a button and then it activates the specific program for that range and then you could implement trim to counteract the steering to the left or right
Im halfway into the video and you probably figured it out but you can make it into a 2 stage process turn off front and back first then turn it back on and turn off left and right currently at 14:25
I was thinking what if you removed the negative and it worked on how I thought it would because when each angle sensor on each side activates then it does a negative on the other side
Early on you had an issue with oscillations. What you need to combat this is something called a PD controller (Although ideally a PID controller would be better)I don't know if this is something even remotely possible to make in Trailmakers, but it might be worth trying to make one as a blueprint to implement in future creations.
As another idea, what if you put down the stabilizer pistons, but hook them up to a distance sensor, and set it so the sensor is set to five foot. Then the entire truck would be pushed up to five feet off the ground, right? Or would that not work because it's uneven ground.... ALTITUDE SENSOR. Sure, you would theoretically need to change it for whatever height you are at, but in theory it should work. Four altitude sensors at each corner to turn on pistons to bring you to the required height above whatever ground level is.
An idea that I had is trying to angle the sensors instead of each sensor being responsible for 2 pistons, causing them to overlap, have each sensor be 45 degrees offset, and only make it responsible for one piston, that way there is no fighting, since they never control eachother's pistons
🤔 I was thinking you could use 4 sensors facing a pendulum, or two sensors facing two pendulums each, at 90°. Then you'll only need if the sensor can see the pendulum to an input to increase on the piston And maybe a reversed set of sensors if they sense the since the pendulum, to decrease the opposing side piston. Anyways that's my 2c... Have a good one, love the vids 👍
Another way you could do it is change the orientation of the angle sensor to a x pattern instead of a +.(this would me you have to use a servo to angle the sensors.) So have one angle sensor pair manage top left and bottom right and one for the other two
I feel like the simplest way to solve this issue is to put them on two different systems. So the left and right would be on its own system with two pistons and front and back would be on its own with two pistons, and they’re both powered by hitting the button 2.
A system of altitude sensors. One on every corner, and one in the center. the one in the center is the average of all 4, so subtracting is your indication. New problem is that the highest corner stays high, so maybe also add and average left to right and front to back. But that would be my startingpoint, take individual altitudes, and subtract the average to get the differences and do math from there
Having two side by side detachable blocks might be a cause for the discrepancy. Like the dynamite drop would it be possible for a magnetic field to stabilize the missile before full launch release?
The easiest, but a bit "cheaty" thing would be one piston in the center and a strong gyro. Still I love this more realistic attempt (so far, currently on minute 12 of the video)
I have 2 other Idea you could using the angle sensors where either you could use them to make some sort of self leveling plane sort of like an auto pilot but a very simple that just keeps the plane going straight when activated this could work for like a challenge like a flying aircraft carrier or something like that I don't where it would be useful but I think it could be cool to have some sort of self level aircraft. You could also create some sort of self leveling missile making it fly straighter for longer reaching further distances or something do with this information what you will but I think you could build some pretty cool things with those angle sensors would love to see more builds with them
you could put the pistons in a diamond shape instead of on the corners so that one set of pistons controls left/rigth and the other controls front/back
multiplayer idea: boats and cruise missiles. build a boat then build a cruise missile (thrust and fly is detachable and in stages. you jet up first then float to your target).
now all you are missing is the rocket on a raised platform where you can input angle to swivel your aim! Then it doesn’t matter if you accidentally repair and face a different direction
Video idea: revisit old videos and give your creations one more evolution. So what I mean here is sometimes one guy will build something that dominates, while the other guy figures what was wrong with his creation early on but you guys don't allow big fixing in the middle of the video, which you shouldn't outside of an evolution video. However, just as you've demonstrated in this video, your creations can be greatly enhanced with the single new idea. Maybe go back to some older videos where some people just got torn up and let the person who got wasted evolve their vehicle and try again. Like when Yeuzi had his jet that kept shooting himself, or professor incognito's craft was just too slow to avoid the human guided missiles... Let those guys try again with an evolution. Another idea, if somebody gets wasted within the first 30 seconds more than twice in a row, they automatically get to try and evolve their craft, something like maybe 5 minutes or 10 minutes or whatever you guys feel like sitting around waiting for them to do. I don't know your schedules obviously. Could be fun!
![NISAMANEE - TADA (ท้าดา) [ OFFICIAL MV ]](http://i.ytimg.com/vi/hcQHEPMmJpU/mqdefault.jpg)
![ค้าง (STILL) - GEMINI [ OFFICIAL MV ]](http://i.ytimg.com/vi/Ftxx3hk5vkU/mqdefault.jpg)
The thing I like about Scrapman is that he always gets there. Like, I'll be sitting here going "slow down the pistons" or "maybe take out the negative?" and he always tries it before he's done. Some builders don't show their process, but he does, including his struggles, and I respect that. Also, at the end of the day, he really knows how to build in Trailmakers. Looking forward to him honing his Screwdrivers skills!
Instead of using pistons on all four sides you can try to make a piston triangle instead of a rectangle. Rectangle having 4 sides can have a side-to-side diagonal wobble issue, but a triangle won't have that issue, it is like how tripods for cameras have three legs only, this prevents the wobble issue. you can use anchor blocks to attach to the ground to fix the drifting and to prevent the vehicle falling due to instability in a tri legged leveling system
The problem with four pistons is because it doesn't know whether a piston is in contact with the ground.
So the negative inputs try to retract a leg to lower that corner, but because it's not on the ground yet, it's lifting it up when it should still be going down to make contact.
@@TheBaldrickkYes, I think most of his problem could have been dealt with by using a manual input to fully extend all pistons and then set the automatic system to retract some of them untill you are level.
That way you would eliminate all the random concacts with ground messing everything up (like the wheels touching the ground sideways while trying to level side to side because the pistons are still too short)
A tripod would have been way more complex to program to level up then a system with four pistons
@@qdaniele97 not really. A tripod will always have all legs in contact.
You adjust the front two for side to side, and then the rear for front to back.
Like all things with legs, if set up properly it won't wobble. Doesn't matter how many legs. The argument that three legs is better than four is absolutely absurd
The simple solution was just to have a tripod system. Two in the front to manage left/right and one in the back to manage front/back. Any three points will always be coplanar, so there won't be any wobble.
11:10 That's the moment Scrapman learned why stabilizers don't use straitgh binary output, but use things like calculus or fuzzy logic
Its just PID :) Multiplication, integrals and derivatives
Multiplayer idea: Teacup race. Each racer has to make a vehicle that will keep a loose piece of cargo on top of it as you race around the desert map. Losing the cargo means you have to stop and respawn the cargo (aka the teacup) before you can continue.
Extreme acceleration in such a race has the obvious drawback of making it harder to keep your cargo on your platform.
I really like this idea
@@seandunkley3055 up?
Good idea, but they need to use a standardized teacup, or else the winning strategy is just 90 degree walls, more like a coffee mug. meaning it never spills unless you almost tip over.
Rather than respawn he cargo though, I think a crane attachment that needs to retrieve it would be better.
@@Nevir202 good thinking
This might be a cool idea Post 1.9 Update with the new Vehicle Collision Dynamics, then the Cargo can't just rip the vehicle apart from the inside of its Cargo Space from just merely innocently rattling around a bit.
19:05ish... ScrapMan: "I wish we had an Emergency Brake."
Incoming 1.9 Update: "Wish granted!" 😘
26:55ish... Scrap, I just got an idea! You spin your entire Missile to spin just the Dynamite... but spinning the Missile throws it off a bit. The Missile never needed the Spin, only the Dynamite Delivery does. Therefore, I propose an idea: Separate the Body of the Missile into 3 Segments with the separation points at just above and below the Missile Body Parts holding the Dynamite. Place 2 Rotating Servos in the Body of the Missile, one on the top and one on the bottom of the Dynamite holding Missile Body, and set them to spin the Dynamite to dispense it while the Top and Bottom Missile Body segments do NOT Spin. Crazy or Crazy like a Fox? :D (Also, wasn't there some kinda Hack where some Part or other was placed beneath the Rotating Servo to trick the Torque into getting Walled there and Gandalf-style "YOU SHALL NOT PASS!!!"...?)
End of Vid: I wonder if any of the new 1.9 Logic will help improve this even further?
Also, I wonder if Yzuei and the others should be scared for future Challenges in how outgunned they're getting in the Logic Force Powers Arms Race...? Yzuei was having a hard enough time trying to face ya in this Challenge with your Truck that did NOT have self leveling capabilities... :P
Simple Solution: Base of the Aiming Mechanism ist build on loose joints. Two 2x1 Joints facing to front, on top two facing to the sides. Now on top Gyrostabalizers set to max. And on top the Launchpad
Yup. Idk why he made it so complex, lol
often it is harder to come up with a simpler solution
Might be easier but it's not a very elegant solution, I'd even say this solution is borderline cheating. Personally I like to build realistic vehicles and mechanisms so I usually stay away from gyros. Good luck finding a real army vehicle that is using your mechanism! Of course gyros do exist in real life but they're like a million times less powerful. People that are fixing all their problems with gyros are just lazy with zero imagination.
@@Alfred-Neuman i dont say, that i didn't like the complicate solution. But i know your point and appriciate it. Also I have to add, not always using a Gyro should be considered cheating as sometimes a "simple" solution is the beginning of getting creative. Sometimes it's better to have a positiv but simple outcome as a negative complex, especially when you are a beginner OR need a q&d solution.
Actual IRL ICBM trucks have a deployable launch pad that folds out the back of the truck and makes the truck immobile while stabilizing the missile and they work in a combination of how scrap man did it and how you described.
21:09 Was going to comment about this when I saw it at 15:31. Glad you figured it out yourself!
I was searching for a comment like this, i was feeling so bad watching that piston activate all the time and him trying to figure out why. I was like "Its the pistoon, its the pistooon"
@@AkutoSuishait was the same for me, i saw him connect it on mistake and wished i could say this to him
The uneven ground is causing issues I think, because It seems to work when the hill is flat (but not level). Try making it balance on 3 points rather than 4. A stool with 3 legs can never wobble, but one with 4 legs always will if the floor is not flat.
i heard there is a proof for non-wobble 4-legged tables on a ground with some limiteto the maximum slope the ground can have. essentially if you rotate the table at the center on that ground, there is an orientation where the table doesn't wobble. though 3 points are obviously easier, but i thought i mention it
@@Akronsusyou would be correct, in some situations.
This obviously isn't able to be implimented compactly into his missile tank.
@@Akronsus Even then, only 3 of the legs are supporting the weight, the other one is in contact with it, yes, but there's barely any weight on it.
It's also what makes rubber/fabric feets on chairs effective, they can squish until the fourth leg is in contact, it's still rocking within the compression of the feets, as it's physically and mathematically impossible to have more than 3 "true" contact points, but so little that you can't feel it.
this is a rather realistic video in terms of both the problem and the solution
i also did a self stabilizing missile launcher and i made by using these steps
1) the pistons all extend until a sensor doesn't detect the ground anymore
2) the gyros stabilize the vehicle by lowering the pistons that need correction
3) fire the distance sensor again but only for a 0.5 sec pulse to just give it some stabilizing buffering height
Well done. I see some comments saying just the launch mechanism but in my opinion using stabilizers to level the whole truck was much cooler. Would like to see you perfect it maybe and use it on different kinds of creations.
you can make it better all you got to do is
step 1: Slap a "sensor" on at EACH "leg" (da pistons dat to da stabilizing) pointing DOWN with less than 1 block DISTANCE (or until they detect the ground at the same time the "leg" touches the ground)
step 2: when the "stabilizing" part is DONE use the "sensors" to CHECK if it DOES NOT detect the ground if so EXTEND the piston until it DOES
step 3: now do the opposite CHECK if it DOES detect the ground if so CONTRACT the piston until it DOES NOT
step 4: do step 2 to 3 again and again until its "balanced"
Note that Im a programmer and been programming for 4+ years so this may work not sure since I cant play Trail Makers (yes... bad PC)
It would be difficult to add with logic just because a loop is hard to make in trailmakers. Needs both several blocks for each step, but it needs a constant input to go around, not sure if it's worth the trouble in trailmakers
@@-aid4084 making a loop (more commonly known as a clock in logic) with logic is easy
@@vexmain2836 sure it's easy in code, not so much in trailmakers
@@-aid4084 based on what I see in sm vids its pretty much the same, easy
gyro stabilizer: Am i a joke to you?
gyro stabilizer 2: What am i? A roach?
Electric boogaloo
For clarification what is the gyro stabiliser 2
@@cogs_and_gears They may have meant a normal gyro vs a gyro stabilizer
The shaking even tho they stabilize You the car will stilul shake a bit
Gyro stabilizers are over rated
Add a distance sensor to a certain distance, 4 on each corner, use an and gate to make the pistons expand until the sensor no longer sees anything
I think it would be really fun to see scrapman try a coding game like "the farmer was replaced", especially because he knows all of this logic and that is the base for a lot of coding! would be very interesting to see his thought processes
Human Resource Machine and 7 Billion Humans would be pretty good initial options I think!
29:35 - Scrapman being Tony Stark lol
Awesome stuff!
I think the issue with the negative input is that they affect both pistons on the opposite side regardless of whether they're contributing or not. If you instead had sensors on the feet of the piston that went into an AND gate you could make sure the pistons never retract when they're not in contact anymore
So the biggest issue I think your facing is having to calculate the extensions of 4 points, if you were to change it to use 3 points instead (back left, back right, front center) you can designate each point to control their specific rotational axis. Front would be responsible for your forward tilt. This will also fix the issue you faced where you are resting on to points and are rocking between them
i dont think thats the issue. the main issue im seeing here is lack of data. its like giving a robot only the altitude off the ground den trying to make it stand up just base of that. the granularity of the information he was working with was jsut way to small for what he wanted to do. shoudl given each piston its own set of sensors.
Thanks scrap the angle sensor stuff really saved me from having to find out how to do it for an automatic build I'm doing
I’m grateful for you, in this sometimes dark and dull world you are a beam of hope for so many people. Thank you Scrapman
This was awesome, Scrapman! The only thought I had was if you were willing to be more manual, you could take advantage of the fact that it doesn't necessarily autoretract when you kill input to separate left-right and front-back auto adjusters two different inputs. Activate the get the tires up pistons, then forward back and then left right. You would need a different cancel button, but you're gonna repair back to neutral anyway so it may not get used much anyway
29:35 gives off Tony stark vibes 😂
In control theory, you created whats called a "bang-bang" feedback controller, where the two options are on/off for control, therefore the name when controlling something loud irl "bang on, bang off".
What you need ideally is a proportional controller, where the control output is proportional to the amount off center. This is the most basic of a "PID controller", which is the standard feedback control loop in engineering
Scrapman- watch me cook guys
Everyone else- ummm wtf? 😂
My brain is about to explode hahah
I absolutely love games with logic (like scrap mechanic or trail makers), because the only time there’s an issue in the logic, isn’t from another factor, there’s a problem that can actually be solved, wind or the ground can’t affect logic, which I just love. It’s very similar to code scripting, like Python, which also runs smoothly if there are no mistakes, similar to logic in games, both scripts and logic take everything literal, with can make them like a big, confusing puzzle.
I like how scrapman has the best ideas and never runs out of them
I feel like the easier way to implement this would be using the pistons as stands to keep the wheels from wobbling and moving the vehicle, and have the auto-correction on the missile launching mechanism. Especially the front to back, instead of trying to adjust the whole car 10 degrees, why not just adjust the launch angle 10 degrees.
Idea: You guys build each other a plane to use. Example: Scrapman makes one for incognito, Incognito makes one for Yzuei, and Yzuei makes one for Scrapman? 👀
*Video idea:* Try and do the same idea (truck launched missles) but in the sea with boats.
Video idea:
Cluster missiles vs a very fast dune buggy thing
Or against a drone/jet
( to make this more balanced I think that you shouldn’t be allowed to Actualy crash into the target, but have to launch your cluster explosives at the target)
(Maybe using a sensor with a large distance that releases the dynamite at the front of the missile would work.)
I also think in this case the biggest issue is, the pistons are retracting, because once you are considered level, the pistons contract because they feel like they are no longer needed due to the angle sensor.
I'm a nearly 14600 days old computer scientist and this is the first time I've genuinely understood XOR. Thx.
You could just say 40yrs …
How could you not understand XOR?
So simple...
If there is no signal going in, there won't be a output
If 1 signal goes in, there will be a 1 output
If 2 or more go in, there won't be a output anymore
The missile's trajectory would be skewed by Magnus Effect, although I'm not sure if it applies in Trailmakers. Depending on how aero is calculated, that might be a contributing factor.
Basically, a wing (via the complex shape) causes air to pass by each side (top/bottom) differently, creating lift.
Magnus Effect refers to how a spinning object can create the same effect without the complex shape. If an object is rotating around an axis not parallel to the windspeed direction, one side will be rotating into the wind, while the other rotates with the wind. These two sides have the same effect as the two sides of a standard wing.
The missile, while flying downward, still has some forward (sideways?) momentum. When it spins, although mainly travelling down (along the same axis as rotation), the sideways drift works with the rotation to create a slight Magnus Force.
Again, I don't know if this is really a factor in the game, but theoretically, you know...?
Love it when Scrapman deepens his logic skills!!
13:41 You could let the middle of your vehicle pivot (on the vertical plane) from the middle, kinda like the opposite plane of travel from the oscillating loader, but you will need peramiters to keep it from twisting past certain points. Then, just level front and back seperatly.
great video, loved seeing the process
The light aim to the left could be more due to the rotational torque just before the release of your explosives from the missile.
So, in regards to the missile skewing with rotational force, (I never realized that was the problem till you pointed it out) I would make vaguely biologically functioning submersibles, and this being before we ever actually had gyros, I made use of helicopter servos. I would run into a similar problem of rolling in strange directions with an odd number of helicopter servos, so you just have to do it like a dual blade copter, two servos in opposite directions.
If you build it properly, the fuselage may not spin at all, but you could still spin the dynamite itself around the fuselage, and then detach after timing, and have the same effect.
This was amazing, Scapman! I want see how you would fix the bugs.
So just a bit of info on control theory and physics that you may be able to apply. Generally the less control operations the better. Meaning less things happening simultaneously the better. I would personally attempt to activate a single piston location at a time. Ie front left, front right, back left, back right. And allow it to balance in all directions at the same time. Just loop through that process indefinitely. I would also only give it a fixed amount of time on each piston, 0.1 seconds and then move on. This will act as a relaxation parameter and will allow the piston to move in the direction of improvement rather than all the way which should eliminate some of the oscillations. The final thing is you are correcting four, which brings nonlinear solutions into the realm of possibilities. If you figure out a tripod configuration you may have better luck stabilizing it as well.
So confusing but so satisfying! Solid effort man!
Next do a vehicle that drives level, adjusting wheel height constantly
I've dealt with this in ScrapMechanic. The way it can work easily is by using 4 angle sensors and rotating them 45 degrees so that they measure the diagonal tilt instead of coronal/saggital tilt.
7:30 oh boy, what's going on now?!- *KABOOM*
Got me dead man
Like a certain spokesperson of a certain game studio would like to say, "It just works".
16 times the logic!!!
I would love to see you revisit this with the next Trail makers update, the new logic blocks, especially the accumulator seem just perfect for this.
scrap i got an idea. pirate ship battles 2 or 3 small ships vs over powered pirate ship. weapons cannons only and a block limit of 100 to 150 for the little ships
For a flat plane you want 3 points of contact, not 4.
With 4 points you need each one to be perfect of you wobble. And programming it will make them fight each other.
With three points of contact(2 in the back and 1 in the front) the two at the back can work on levelling left/right with the front pivoting to conform, then the front one can adjust the front/back level with the back ones pivoting to conform
Easy use a toggleable gyro-stabiliser in each direction on max power
Have two sensors per side and they only deal with that side of the vehicle. Two pitch on the right side, two on the left, two roll on the front and two on the back. That should solve some problems
One potential solution would be to have a set of pistons for roll and a second set for pitch.
This would prevent the two axis from sending conflicting inputs to the same piston.
Scrapman: "I'm not a logic person"
Also Scrapman:
Suggestion: *[TLDR available]* Use pistons to clip in X number of Dynamite, and have each of them connected to a hinge and that connected to a coupler (disconnecter) so the hinge folds the dynamites upwards, and when you need to release it, you first fling the dynamite with the hinge going down (Adds momentum) and on a short delay, seperate them, this will make a good non-spin and high accuracy missile... Might be a hassle to make but maybe you can also do a TrailMakers MM challenge, like someone makes a missile and someone makes an armored bunker, and you can have Yzuei make the missile and you make the bunker? And the challenge is that the team that takes the least shots to destroy the bunker (destruction can mean anything: Maybe putting a buncha dynamite in the center that is covered up by the bunker and if that explodes, then you're out???)
TLDR: Instead of spinning, flick the dynamite outwards with a hinge...
The chickens get no mercy from Scrapman💀
Winner winner chicken dinner! 😎
@@JoshuaGanoTyraxLightning Spoken like a true Florida man
@@guywithincrediblylongusername It's funny to hear that while i'm actually living in Washington State. :)
Try cycling responsibility, level length wise => wait => level width wise => wait => repeat.
7:30 DELAYED ANNIHILATION!!
One issue was that with the negatives inputs included, whenever one leg is extending, the opposite corner leg will always contract (no matter if its front to back or side to side), creating the wobble. A three leg system would have been easier to manage because it would never tilt and the position of each leg is directly corresponding to the motion of the truck.
To get the wheels off the ground, an easy solution would be to just add some extra pistons which are active no matter what hwen the stabilization is on
I'm going to assume to you saw my comment on that video for this idea! :P
It would've been easier to use the pistons to auto-correct the left and right angle only and adjust the front and back angle through the missile angle directly.
To have it automatic both ways with pistons, the easy way would be to have a set of pistons for front and back angle only, and use the other set of piston for left and right angle only. Your system is pretty neat though, and it should have more range of motion for extreme angles. Good job!
Just put the rocket on a gyroscopic platform with a low centre of mass or even a gyro stabilizer
(in middle of the video) I thought of
Each of the legs having separate control diagonally X not left right and front back 🔳 so they don't fight together but cooperate 🤔 (endedit
Basically the sensor needs to be not left right but ↙️↗️↖️↘️
It's should be technically the same thing with logic just 45 degrees
And each need 1 piston to lift wheels of ground separately.
27:20
Why do you never do the same logic but with slower and only 0.3 degree range (instead of 2) that way you won't overcompensate which is often caused by overshooting the range of degrees thanks to momentum or and delay of deactivation
Enjoyed the video hope you read my tips 👍
I think the solution would be for one corner of the pistons having 4 angle sensors, so every corner will having own x and y axis. 2 sensors will extending the pistons and the 2 sensors will retracting the pistons (for example the back left corner will having for X axis the one sensor that is 90 direction and 179 width will retract the pistons and the one with -90 will extend the pistons)
A video idea-
You could build air defence system.
Something that could help would be to rotate where the 4 piston sets are so that there's one pison on the front, one on the right, one for the back, and one on the left. That way there's no doubleing in the programming. Each angle sensor will only output to a single piston.
Thinking about this now, though if the vehicle is tilted forward and only a front center piston extends, the vehicle will only be supported by 2 pistons, becoming unstable...
Idea to make things easy: Always make a simulation of what are you trying to do mathematically in some programming language first, so you can correct the errors in there, and also, use the new gates
I had the exact problem with my missle launcher but i went with a Missle Bay that rotates around the longitudinal every Missle Pod lat angle alinged according to the sensors. The advantage is that you can fire while on the move
Hello, I've made a Harrier at trailmakers but I'm embarrassed to upload a video on TH-cam so I posted it in the gallery (the design is called "Harrier G3 (no landing gear)" it's a nice harrier :)
try to make many different launch programs such as 100m 200m 300m 400m 500m so that you just push a button and then it activates the specific program for that range and then you could implement trim to counteract the steering to the left or right
GG. I think the reason the gyro is changing the course of the rocket is because it is not acting thru the centre of mass.
Give this missile truck a rotating turret. Also for your next plane build, use hover pads as landing gear.
well done! (though a triangle of pistons would've been better than a rectangular layout probably)
also nice vid
been here since the multiplayer mondays when camodogaming was in them, love the vids!
Im halfway into the video and you probably figured it out but you can make it into a 2 stage process turn off front and back first then turn it back on and turn off left and right currently at 14:25
I was thinking what if you removed the negative and it worked on how I thought it would because when each angle sensor on each side activates then it does a negative on the other side
I love how we were thinking alike on most of it
i think you could of done it with distance sensors but it is fun to watch you find solutions to issues
Scrapman becoming Tony Stark arc is amazing and I'm here for it
Great content as always, Thanks
2:45
Chicken Killed; Snipe-Bomb!
Having separate pistons for forward/back and left/right stabilization could make the logic a lot simpler
Ive been waiting for him to do this ever since the cluster missile video
Early on you had an issue with oscillations. What you need to combat this is something called a PD controller (Although ideally a PID controller would be better)I don't know if this is something even remotely possible to make in Trailmakers, but it might be worth trying to make one as a blueprint to implement in future creations.
As another idea, what if you put down the stabilizer pistons, but hook them up to a distance sensor, and set it so the sensor is set to five foot. Then the entire truck would be pushed up to five feet off the ground, right? Or would that not work because it's uneven ground.... ALTITUDE SENSOR. Sure, you would theoretically need to change it for whatever height you are at, but in theory it should work. Four altitude sensors at each corner to turn on pistons to bring you to the required height above whatever ground level is.
you can probably try to make a simple PID in scrap mechanic. would be a fun video
Lookin forward to another video
I be staying up late to watch these fr
An idea that I had is trying to angle the sensors
instead of each sensor being responsible for 2 pistons, causing them to overlap, have each sensor be 45 degrees offset, and only make it responsible for one piston, that way there is no fighting, since they never control eachother's pistons
🤔 I was thinking you could use 4 sensors facing a pendulum, or two sensors facing two pendulums each, at 90°.
Then you'll only need if the sensor can see the pendulum to an input to increase on the piston
And maybe a reversed set of sensors if they sense the since the pendulum, to decrease the opposing side piston.
Anyways that's my 2c... Have a good one, love the vids 👍
There is a new trailer for the 1.9 update with alot more logic and aerodynamic blocks
Another way you could do it is change the orientation of the angle sensor to a x pattern instead of a +.(this would me you have to use a servo to angle the sensors.) So have one angle sensor pair manage top left and bottom right and one for the other two
I feel like the simplest way to solve this issue is to put them on two different systems. So the left and right would be on its own system with two pistons and front and back would be on its own with two pistons, and they’re both powered by hitting the button 2.
This is Scrapman's equivalent of coming up with the perfect comeback two weeks later
A system of altitude sensors.
One on every corner, and one in the center. the one in the center is the average of all 4, so subtracting is your indication. New problem is that the highest corner stays high, so maybe also add and average left to right and front to back.
But that would be my startingpoint, take individual altitudes, and subtract the average to get the differences and do math from there
chilling in an ac room laying down on the room phone on charging a quilt on top of me and hot noodles plus some monster energy and scrap mans video
Average 14 year old
Hell yeah
Some punctuation would do a lot in this mess of a paragraph
the room phone
charging a quilt
Having two side by side detachable blocks might be a cause for the discrepancy. Like the dynamite drop would it be possible for a magnetic field to stabilize the missile before full launch release?
The easiest, but a bit "cheaty" thing would be one piston in the center and a strong gyro.
Still I love this more realistic attempt (so far, currently on minute 12 of the video)
I have 2 other Idea you could using the angle sensors where either you could use them to make some sort of self leveling plane sort of like an auto pilot but a very simple that just keeps the plane going straight when activated this could work for like a challenge like a flying aircraft carrier or something like that I don't where it would be useful but I think it could be cool to have some sort of self level aircraft. You could also create some sort of self leveling missile making it fly straighter for longer reaching further distances or something do with this information what you will but I think you could build some pretty cool things with those angle sensors would love to see more builds with them
I bet it would've been easier to have done this with 3 legs with only 1 in the front, but props for getting a 4 legged system to not keep rocking
you could put the pistons in a diamond shape instead of on the corners so that one set of pistons controls left/rigth and the other controls front/back
multiplayer idea: boats and cruise missiles. build a boat then build a cruise missile (thrust and fly is detachable and in stages. you jet up first then float to your target).
now all you are missing is the rocket on a raised platform where you can input angle to swivel your aim! Then it doesn’t matter if you accidentally repair and face a different direction
Video idea: revisit old videos and give your creations one more evolution. So what I mean here is sometimes one guy will build something that dominates, while the other guy figures what was wrong with his creation early on but you guys don't allow big fixing in the middle of the video, which you shouldn't outside of an evolution video. However, just as you've demonstrated in this video, your creations can be greatly enhanced with the single new idea. Maybe go back to some older videos where some people just got torn up and let the person who got wasted evolve their vehicle and try again. Like when Yeuzi had his jet that kept shooting himself, or professor incognito's craft was just too slow to avoid the human guided missiles... Let those guys try again with an evolution. Another idea, if somebody gets wasted within the first 30 seconds more than twice in a row, they automatically get to try and evolve their craft, something like maybe 5 minutes or 10 minutes or whatever you guys feel like sitting around waiting for them to do. I don't know your schedules obviously. Could be fun!