Hello, I'm new to this forum and Arduino. I need someone can help me design (or give me idea) what part I need for my purpose.
Here I want: I want to use Arduino system to control a 12v hydraulic pump and hydraulic valve, so they can control a double acting cylinder go up and down when sensors send the signal.
Here is my idea about power leg skeleton: I try to build a hydraulic power leg skeleton so that it can give the user more power to handle with weights. When the leg of user moving, the hydraulic skeleton moving following and gives user some extra power. That is my idea, but I really need help with the question: what kind of sensors I can use for this purpose. I'm almost done with the mechanic design. I need someone help me for the controller.
hydraulic ? Sounds dangerous . Shouldn't you prototype it with pneumatics first to develop the software and then build it with hydraulics after you have the software working ?
I like ur ideal about flex sensor. I want sensor that can describe direction of force(up or down,so it is simple), how quick the force accelerate/decrease.
raschemmel:
hydraulic ? Sounds dangerous . Shouldn't you prototype it with pneumatics first to develop the software and then build it with hydraulics after you have the software working ?
I never think why it is more dangerous than hydraulic. Pneumatic is more simple,in this case, Ecuador only controls the valve,but the needing for air supply and the sound of air pump really annoy. That is my thought. Please correct me if I am wrong
ketoi12345678:
I like ur ideal about flex sensor. I want sensor that can describe direction of force(up or down,so it is simple), how quick the force accelerate/decrease.
Those flex sensors only work in one direction of flex though, which might limit that idea.
I think a pressure sensor would probably be more suitable than a flex sensor. If you can identify a few contact points where pressure-sensitive pads can be placed then you could design your controller to move the exoskeleton away to relieve the contact pressure. I imagine it would be extremely difficult to make something that worked well enough to be usable in the real world, but you ought to be able to get something working well enough to demonstrate the concept.
ketoi12345678:
When the leg of user moving, the hydraulic skeleton moving following and gives user some extra power. That is my idea, but I really need help with the question: what kind of sensors I can use for this purpose. I'm almost done with the mechanic design.
I can see this being a very challenging project.
It would help if you show us a sketch of the mechanical design.
How is it linked to the user's leg?
How is the load to be shared between the user's leg and the mechanical device?
Are you thinking of sensors that detect nerve impulses or muscle contractions?
Or are you thinking of some mechanical sensor that detects when the movement of the user's leg is ahead of the movement of the mechanical device - for example the angle of the knee joint of the user might be different from the angle of the mechanical knee so it must move to match the human knee.
PeterH:
I think a pressure sensor would probably be more suitable than a flex sensor. If you can identify a few contact points where pressure-sensitive pads can be placed then you could design your controller to move the exoskeleton away to relieve the contact pressure. I imagine it would be extremely difficult to make something that worked well enough to be usable in the real world, but you ought to be able to get something working well enough to demonstrate the concept.
Thanks ur idea.
How about if we combine flex sensor and rotation sensor. So flex sensor will describe how much force,and rotation senseor will describe the position.
Or can we use MINIMU-9 V2 GYRO, ACCELEROMETER AND COMPASShttps://www.robotics.org.za/index.php?route=product/product&path=57_129&product_id=204
I can't help but think, and I know it's too late to say this, that the control system should have been designed in rather than trying to fit it to an almost complete mechanical design.
This is surely where the discipline of mechatronics- which my daughter plans to study at uni from next year- comes from: it needs an integrated approach of mech, elec, control etc etc to make a project like this a real success.
Let me explain:
Flex sensor (A): describes how much fore at the hip
Rotation sensor (A): describes direction of force (up/down)
Combination of the two sensors will send signal to gear box to control the upper leg (from knee to hip) move up/down, and the speed (slow/fast)
Flex sensor (B) and rotation (B) (I forget to draw it =(): to control the lower leg (from knee to hip), using same idea.
These sensors will be mount on human body, the skeleton will mount sideway, like this:
Here I am not sure/ need help:
with the flex sensor, can it be bended forward and backward? 2 ways to bend or just 1 way. If it can be bended 2 ways, so I don't need the rotation sensor
For example: flex sensor's vertical with ground, bend it forward an angle 30 degree, it will give some value. Then bend back to an angle 60 degree, it will give some value or not?
I plan to use hydraulic cylinder in this system, but it really hard to find info about using arduino to control hydraulic flow valve. I even cannot find a 12v hydraulic flow valve (using to control the speed of hydraulic cylinder)
Just watch the movie: Edge of Tomorrow, cool power skeleton. I want to build a prototype, so please help me with these questions. Any idea, thought, comment (good/bad) are welcome.
What is my goal: build an armor like Iron man, but for now I just want a power leg skeleton to help users have more power for heavy jobs, but most my goal is to help handicapped people can walk by themselves. I want you guys, be a partner. Therefore, please help me gain knowledge from you. Thanks for spend time on reading.
I suggest you start with a small model of one joint - perhaps two pieces of wood with a screw connecting them. Use a potentiometer to detect the angle - fix the body of the potentiometer to one arm of the joint and the "knob" to the other arm.
Then use a cheap servo to make a similar joint move in synch when you bend the first joint.
Don't even think of anything more elaborate until that works really well.
And maybe the first thing to decide is how you will know it is working really well.
One of the problems I foresee is that if the mechanical device is sufficiently linked to the human leg so as to carry weight that would otherwise be carried by the human leg how can the leg move within the device to signal that a change of angle is needed?
Or if the leg is free to move within the device how can the device take the load off the leg?
hydraulic ? Sounds dangerous . Shouldn't you prototype it with pneumatics first to develop the software and then build it with hydraulics after you have the software working ?
Even more dangerous as the working fluid is compressible.
A search for things like 'mechanical exoskeleton" should produce numerous previous projects. Pots can be used on most joints to provide position indication. You might get some ideas from the below simple robotic arm setup.
I can't remember whether it's from one of the Iron Man movies or RoboCop, but I remember a scene when they were demonstrating a suit and it malfunctioned, ripping the driver's lower torso off and kill him. Think that was Iron Man.
EDIT: It was from IronMan 2. Oh yeah, and according Justin Hammer, the test pilot survived. Lol.
Robin2:
I suggest you start with a small model of one joint - perhaps two pieces of wood with a screw connecting them. Use a potentiometer to detect the angle - fix the body of the potentiometer to one arm of the joint and the "knob" to the other arm.
Then use a cheap servo to make a similar joint move in synch when you bend the first joint. I don't get it yet, could u please explain?
One of the problems I foresee is that if the mechanical device is sufficiently linked to the human leg so as to carry weight that would otherwise be carried by the human leg how can the leg move within the device to signal that a change of angle is needed?
Or if the leg is free to move within the device how can the device take the load off the leg?
First, off, I saw nothing in the video that gives even a hint at how the body controls the device and, in view of the $millions they are obviously spending on its development I'm not surprised. I suspect its computer is at least as powerful as the most powerful laptop you could find.
All I meant by the stuff you highlighted is to make two simple joints and with simple components develop some code so that moving one joint causes the other to mimic its movement exactly. If you don't understand what I meant by the use of the potentiometer and servo let me know. But I assume you do since it is very trivial compared to the project you are thinking of.
You should search for some of the stuff that Dean Kamen (designer of the Segway) has done to get some idea of the complexity of the feedback systems needed to mimic human behaviour.
