Stewart Platform with Force Control

[@julianandresgomezgomez7264]

Amazing work! Congratulations.

[@gaurishgangwar]

Thumbs up for the software interface 👍

[@ashutoshdayal2609]

Thanks.

[@NurJamiludin]

I'am interested in stewart platform And try to make it my self Can you share me the detail design of your work, where can i find to purchase the spherical joint. ?

[@ArunUdai]

You can mail us to [e-mail]

[@BillDWP]

Awesome, thank you for sharing.

[@정문원-o9q]

It's Cool! Hello. I'm researching on the Stewart Platfrom. Do you have paper related to this work? If so, could you tell me the title? Thank you.

[@ArunUdai]

Yes: One of them is here www.sciencedirect.com/science/article/abs/pii/S0736584518303296

[@ArunUdai]

Another one is here ismej.iitd.ac.in/mechanicsanddesign/index.php/announce/download/6

[@shuha3216]

Very cool force controll!

[@chadgaspard]

Could a Raspberry Pi be used as an open source means to fetch data refs from any flight sim that provides flight control force data refs and utilize those data refs to drive a force feedback control yoke and rudder pedals without the need to download a third party sim specific plugin? Maybe the Stewart Platform, open gl, and/or python could accomplish this?

[@elsnowman123]

I know this is a little late, but I'll describe how you can do this. Raspberry Pi would be a little overkill for this, and likely more difficult. You'll want to use an ATMega32u4 microcontroller (Arudino Pro Micro development board). This will let you easily use the arduino library to implement USB communication, which should be easier than bare-metal USB (which I've done, I don't recommend it unless you're trying to learn embedded development specifically). The 32u4 will be used as a USB device, receiving force commands from the PC and sending axis data to the PC. This means the 32u4 will have to implement the USB HID (Human-interface device) class in addition to the HID PID (Physical Interface Device) class, which is more complicated than the HID. The HID report is used for sending rudder and yoke data to the PC and the PID report is used for receiving force commands. On the host side, you'll need to do three things: Receive flight data from the simulation, calculate the forces the yoke is experiencing, and send force commands to the 32u4 over USB. C# is probably your best bet for doing this, although python could be used as well (any language could be used but these two will be easiest). Receiving flight data from the simulation can be a little tricky depending on the simulation, so its up to you to figure out how to receive the data. Communicating with a running process is called Inter-Process Communication (IPC) and it's accomplished in a few standard ways, so I'd recommend reading on that. It's up to the simulation to decide how they want to do it, so some do it in different ways. For example, in the racing simulator Assetto Corsa, they provide a port for other processes to connect to in order to receive real-time data from the game. Calculating the forces is up to you since its physics stuff. Sending the forces to the device will require you to use DirectX DirectInput in C#, so read up on that. You can also read on how to receive data from a USB HID device. Here's a useful link for sending force commands: stackoverflow.com/questions/23914486/send-feedback-effect-to-joystick-from-net Now, to actually create forces on the pedals and yoke, you'll need some type of motor or actuator. You could use a PMDC motor with pulleys to create the forces (probably the easiest) or use solenoids, etc. Up to you. You'll also need to read the position of the yoke and pedals, which can be done with a number of rotary encoders connected to the different axes. You'll need some sort of separate power supply to power the motors. You could use some server power supplies for that. I'm using the HP-DPS800GB power supply for a force-feedback sim racing system I'm making. There's a guy on youtube I've talked to before that made his own force-feedback yoke. I'd recommend watching his videos if you're curious how he does the force calculations from the flight sim data. Good luck!

[@lukmannugraha9846]

What is the control method applied for this project? Is this PID ?

[@ArunUdai]

YES

[@elijahantony4300]

Hi, is there a name for those ball joints?

[@ArunUdai]

Yes, the one used here are Rod End bearings.

[@liutaoshi150]

This is really cool, I'm currently working on a racing simulator related equipment, this is just right for dynamic seats, can I ask you more questions?

[@ArunUdai]

Sure

[@liutaoshi150]

@@ArunUdai Can you leave me an email? I will contact you by mail

[@liutaoshi150]

My Email is [e-mail]

[@ArunUdai]

@@liutaoshi150 [e-mail]

[@liutaoshi150]

@@ArunUdai I sent you an email to [e-mail]. Looking forward to your reply

[@ukissrulez]

Whats the use of this?

[@ehabhti]

Great work . But this motor have realtime position feedback ?

[@ashutoshdayal2609]

Yes

[@dylanreynolds47]

Hi. Do you have a resource for the kinematics? I have done my own mathematical model, but wanted something to compare to. Thanks.

[@ArunUdai]

You can use ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=6778142 as a reference.

[@ArunUdai]

Please see this cdn.intechopen.com/pdfs/34392.pdf

[@DassHunters]

Hello. I really like this design. I'm working on a research project with regards to off-shore wind turbines. Have you made any publications on this platform, if so I would be very interested. Thank you.

[@ArunUdai]

Karsten Kopperstad we are working on. We will put the link here as soon as it gets published. Thanks.

[@pushparaj1429]

@@ArunUdai Hello Sir, Is it published? I am working on a project in that also I have to control the system in 6D. So interested to learn. And also Can you share the details of the UI design/Simulation work?

[@ArunUdai]

@@pushparaj1429 www.sciencedirect.com/science/article/pii/S0736584518303296

[@ArunUdai]

www.sciencedirect.com/science/article/pii/S0736584518303296

[@moathmomani2887]

its really nice model, i want to build one for my self so can you help me with resources for the mathematics ?

[@ArunUdai]

The inverse kinematics that is used to do this is published. You may look at my papers at www.dayalsoft.com

[@vincentzhang301]

Hi! It's really cool!Can you tell me what this simulation software is and can this platform be controled by the legs length?Thanks!

[@ArunUdai]

The software was developed by us using Java Open GL. Yes, normally Stewart Gough platforms are controlled by controlling its leg lengths.

[@vincentzhang301]

Thank you! I'm an undergraduate and i want to use this platform to make a neck of a robot but i don't know how to simulate it.I can never thank you enough if you can provide this software with me.

[@WLDLNG]

This is really interesting. How did you manage to achieve force control with position controlled motors?

[@ArunUdai]

It was through current limiting and thereby limiting its torque.

[@akshayjose6555]

Hi i'm so interested in this project, what is the current status any updates.. ?

[@ArunUdai]

Akshay Jose The project is now with Nugenix, Pune. www.nugenix.co.in

[@elroimacedo5255]

Ficou top de mais...

[@ArunUdai]

Muito obrigado pela apreciação

[@Soscha55]

Hi. This is really cool. 👍 How did you control the 6 Dynamixels? With the robotis controller or is there an other option? And which motors did you use? I am working on a similar project an try to figure out which are the best servos for this project.

[@ArunUdai]

Shade Red Directly through its API using usb2dynamixel adapter. We used MX-64T for this work, however any motor is suitable depending on the scale.

[@Jschic1989]

Awesome, congrats for your work! Just one question, what Dynamixel motor references did you use?

[@ArunUdai]

MX-64T

[@Jschic1989]

Arun Dayal Udai and what controller board did you use to simultaneously control the 6 servomotors?

[@ArunUdai]

@@Jschic1989 Dynamixel actuators do not require any controller board. It uses Serial communication from PC and all actuators are connected in a single serial bus. And a 12V supply is required separately. Check with robotis.us website for full details.

[@Jschic1989]

Thanks for your answer!

[@nagamanigi]

Hello. This seems to be very interesting. I am curious to know, how did you do the force control with Dynamixel?. Since it uses high gear ratio. Are you using any other sensor to measure the external force?.

[@ArunUdai]

Hi, thanks for watching so carefully. Dynamixel MX-64T supports torque commanding using current control. However, it is not so precise, probably due to the high gear ratio as you have pointed out. One can have the robot Jacobian and can relate the end-effector forces to the joint torques. We are working in open loop without any sensor. Having a sensor will enable us to achieve better precision using a feedback mechanism.

[@nagamanigi]

Thank you! Sir for your reply.

[@Siya--2005]

Is your friend Abhishek. He is my friend....

[@kameshwarsingh4979]

Theme Man: are you telling me. But this video is awesome...👌