Пікірлер

• 
  @MackjSmith-wr4jjКүн бұрын

  This on on the top after the other befroe it the somatice custodians robot's coming soon near you

• 
  @crossingtimezones87322 күн бұрын

  It's not clear to me why q has 3 DOFs even though it has 4 parameters. Can you please elaborate on this?

• 
  @RedragonKid15 күн бұрын

  Very helpful! Thanks!

• 
  @jiyanbaranbukun18 күн бұрын

  I am a complete beginner in ROS and this is an amazing content. Thanks a lot for your effort.

• 
  @roger734127 күн бұрын

  I have developed or directed the development of several large-scale multibody dynamics programs and have never used the Denavit and Hartenberg method, which forces kinematicians to use unconventional methods to fit their models to DH limitations. It is much simpler to split shape matrices into sequences of primitive translations, tx, ty, tz, and rotations rx, ry, rz, in any order. And it is much simpler to represent joints as sequences of primitive translations, tx, ty, tz, and rotations rx, ry, rz, in any order. Just string their spatial displacement transformation matrices together in any desired order, and extract constraints, velocities, and accelerations as desired.

• 
  @roger734127 күн бұрын

  Are robots left-handed? Why are we using left-handed coordinate systems?

• 
  @roger734127 күн бұрын

  The degrees of freedom (DOF) of any articulated mechanism is its mobility number, such as 0 or 1 or 2 or, etc. DOF is also the sum of all constrained joint mobility numbers. A kinematically locked joint has zero mobility or zero equivalent mechanical advantage, and a dextrous joint has nearly unity mobility and high equivalent mechanical advantage. The ith joint's mobility number may be obtained from the square of the norm of the ith row of the constraint Jacobian matrix's orthonormal null-space matrix. The number of columns in this matrix is DOF, as is its norm. The sum of squares of all row norms of this matrix is also DOF, which is why the sum of all joint mobility numbers is DOF. Technically a joint's mobility number is the square of the direction cosine between it and the mechanism's differential constraint manifold. Maximum motion and mobility is obtained when the joint is closely aligned with the manifold's surface, and minimum mobility when it is orthogonal to or nearly orthogonal to the surface. The most reliable way to get the correct DOF, which can also fail, is to apply Gaussian Elimination with Complete Pivoting (GECP) to a very accurate Jacobian matrix associated with the articulated mechanism's kinematic constraint equations. If the model has n variables and the Jacobian matrix has rank r, then DOF=n-r of the variables are independent. The wrong rank r and thus the wrong DOF may happen when mechanisms are over-constrained and some or all of the joints must be perfectly aligned to get any motion at all. At each stage of matrix factorization there is a remaining residual Schur complement matrix. When the norm of this matrix is zero or falls below some predetermined constant, the Schur complement matrix is considered to be zero and factorization stops. Matrix rank r is then the dimension of the resulting invertible factor matrices, not the number of rows or columns of the Jacobian matrix. Problems occur when the Jacobian matrix is noisy or in error for whatever reasons, and the noise or errors contaminate the remaining Schur complement matrix, causing the algorithm to incorrectly increase rank r and decrease DOF. Get the wrong DOF and the mechanism model will not perform as expected. A good example is a 6R Bricard mechanism with a 6 by 6 Jacobian matrix. When the six joints are laid out correctly and precisely the mechanism will have one DOF and articulate. Thus the matrix rank r must be 5, so DOF=6-5=1. But even tiny amounts of constraint error will lead to tiny amounts of error in the Jacobian matrix, and GECP will return a rank r=6 and DOF=0. Before factoring the Jacobian matrix to determine matrix rank and model DOF, the constraint violations must be minimized to the extent of available precision, and hopefully that will be enough for GECP to get the correct matrix rank and DOF. Incidentally, see what the Chebychev-Grübler-Kutzbach criterion is for this mechanism and ask yourself which method is more likely to be more accurate most of the time. See "Chebychev-Grübler-Kutzbach's criterion for mobility calculation of multi-loop mechanisms revisited via theory of linear transformations" by Grigore Gogu

• 
  @roger734127 күн бұрын

  What if a robot had learned what a toddler learned in her mother's kitchen pulling and pushing on the knobs and handles of doors and drawers? We sense where and how to push and pull and twist on objects to make them do what we expect. Imagine a robot with template kinematic and dynamic models of objects surrounding it embedded in its brain with the ability to probe mechanism mobility anywhere and in any direction and pick what works best instead of always by trial and error. A mobility number is like a perceived mechanical advantage. If the mobility number, which is the square of a direction cosine relative to the differential constraint manifold surface, is close to unity, the door will open easily, but if it is zero or close to zero, the door will not budge.

• 
  @roger734127 күн бұрын

  It's all in how one interprets it. Forty years ago I developed my own matrix-based spatial algebra notational convention, similar to Featherstone's, and used it to develop a general purpose, 3D rigid multibody dynamics modeling and simulation program for a fortune 500 company. Let there be one inertial frame 0, and believe me, the numerical processing is much simpler and more efficient when all vector coordinates are expressed in the same frame 0 and all bound vectors are moved to the same frame 0 origin. Imagine a spiky ball of hundreds of bound vectors all intersecting at a common point with all those free vectors floating around that you can't seem to lay a finger on. It's sort of like protons and neutrons sitting in the middle of an atom with a cloud of evasive electrons that don't seem to be anywhere at the same time.

• 
  @tumsfestival802728 күн бұрын

  But can it unstop a toilet?

• 
  @muhammadmohidfarooq456Ай бұрын

  Can i get the code??

• 
  @Rami7605Ай бұрын

  thanks for this tutorial

• 
  @Rami7605Ай бұрын

  amazing coding skills

• 
  @Rami7605Ай бұрын

  thanks for this wonderful tutorial

• 
  @sultanhangaАй бұрын

  Amazing video, thanks

• 
  @sultanhangaАй бұрын

  does this work if there is only ubunto on my laptop

• 
  @mecharithm-roboticsАй бұрын

  yes, it will.

• 
  @sultanhangaАй бұрын

  ​@@mecharithm-robotics thank you it worked for me, although I had to take several steps.

• 
  @ali-k643Ай бұрын

  imagine if you are sitting in the office next door, this robot comes to you say hello and then start spraying all over your face😂

• 
  @moldo800Ай бұрын

  Great video, Great explanation esepecially on Grubler's formula ps. I love it when Salma Hayek explains Mechanisms :D

• 
  @A.Hisham86Ай бұрын

  I didn't understand how a rotation and a translation in the same time compose one degree of freedom?

• 
  @mecharithm-roboticsАй бұрын

  That's great question. Because in that specific joint, the motions are not independent of each other and happen at the same time thus having only 1DOF.

• 
  @user-li6pz9vw4uАй бұрын

  This is fukcing linux

• 
  @alexcartwright8197Ай бұрын

  This series has been fantastic - I tried working through the Modern Robotics textbook on my own but it has been difficult for it to really click. Your videos have been amazing, thank you. Will you continue this series?

• 
  @mecharithm-roboticsАй бұрын

  Hi Alex, thanks for your feedback. Dr. Madi continued this series in her Github: github.com/madibabaiasl/modern-robotics-course/wiki and she said that she has plans to make the video versions available in the near future. But the current github can guide you to really "feel" the math in practice as well (it has implementation on a robot arm both in simulation and reality).

• 
  @alexcartwright8197Ай бұрын

  @@mecharithm-robotics Thank you very much!

• 
  @s.denizkaradag6619Ай бұрын

  I am just installing the program and everything with your videos, I have done every step but I dont have the catkin_tools_prebuild document is that would be a problem?

• 
  @AhmedAdel-kg5qnАй бұрын

  Hello Doctor, why did you not post a series on a mechatronics course or film lectures on the mechatronics curriculum during an explanation at the college?

• 
  @du5kcrumble209Ай бұрын

  What about the code brown protocol? 💩

• 
  @sorenhubble6531Ай бұрын

  now add SLAM

• 
  @Michael-np7geАй бұрын

  49:45 can use print(str(time.to_sec()) + " seconds"). The .to_sec() method converts the Time object to a float of seconds then str() casts the float to a string for concatenation.

• 
  @Michael-np7geАй бұрын

  For getting the Time variable into a float At 49:45 we can use the Time method .to_sec(), then cast the float to a string so we can concatenate with " seconds. e.g. print(str(time.to_sec()) + " seconds")

• 
  @dehanwiraputra2241Ай бұрын

  the one robot tricking the other one to return to the start was absolutely diabolical

• 
  @aslanmonahov6238Ай бұрын

  21st century and they are still dumbass as fuck

• 
  @khantalks02 ай бұрын

  Dr I want to join your lab. Any open position for master degree

• 
  @mecharithm-roboticsАй бұрын

  Dr. Madi's not overseeing comments here. If you are looking for open positions, please contact her directly via her school email.

• 
  @khantalks02 ай бұрын

  Amazing work Dr

• 
  @ClericHeretic2 ай бұрын

  Amazin course!

• 
  @QueenSorrow51502 ай бұрын

  LoL.... That's what they believe a bathroom looks like when someone cleans it. I saw a lot of mistakes.

• 
  @muhammadtehreem41462 ай бұрын

  The norm of the linear velocity component of the Twist vector is 4, not 1.4142 at 19:57. Could you please guide?

• 
  @Dr.ShrinathaRKatti2 ай бұрын

  catkin init and catkin build commands are not working. it says 'command not found'. Pls help

• 
  @monisha1922 ай бұрын

  can you make the same tutorial in ROS2 plz

• 
  @mecharithm-robotics2 ай бұрын

  We are currently working on that.

• 
  @monisha1922 ай бұрын

  May i know when that will be released, cuz I'll be the first one to watch it 🙏🏼😁

• 
  @mecharithm-robotics2 ай бұрын

  @@monisha192 then you will be the first one to be notified :D in the meantime, you can go through these series of lessons where Dr. Madi Babaiasl from SLU has used ROS2 to teach hands-on robotics: github.com/madibabaiasl/modern-robotics-course/wiki This will give you the first impressions of ROS2 until the complete tutorials are out there. Note that all lessons can be done in simulation only and no need for a physical robot arm.

• 
  @monisha1922 ай бұрын

  @@mecharithm-robotics thank you very much

• 
  @lunalunita9752 ай бұрын

  Dude thank you so much, I was thinking I had to buy another ssd.

• 
  @dhirenvairagade65472 ай бұрын

  Unable to locate package os-uninstaller

• 
  @artinmokhtariha41873 ай бұрын

  Thanks for your high quality tutorials It was pretty perfect 🥂👌🏻

• 
  @leandrogcosta3 ай бұрын

  Astounding

• 
  @DIYD3 ай бұрын

  This is kind of old isn't it? We had a toy like this in the early 90s from the Air and Space Museum 🤷🏾‍♂️

• 
  @cfofana45023 ай бұрын

  To the guys who have built it, you are destroying people's jobs for your only benefit !

• 
  @leandrogcosta3 ай бұрын

  Wow

• 
  @leandrogcosta3 ай бұрын

  Wow

• 
  @leandrogcosta3 ай бұрын

  Wow

• 
  @anzone19573 ай бұрын

  Where to buy this robot?how much its cost?

• 
  @benimadimeva3 ай бұрын

  Dünyanın en ikonik videosu. :)

• 
  @alexfish77923 ай бұрын

  These videos are gold.

• 
  @mecharithm-robotics3 ай бұрын

  Thanks Alex. If you like to learn the same level of robotics but with some hands-on experiments using ROS2, you can also follow these lessons: github.com/madibabaiasl/modern-robotics-course/wiki they are basically the extended version of the videos but with more practice. note that although Dr. Madi used a physical robot in the lessons, she always explain that they can be done in the simulation as well (she shows how to do that in the lessons).

• 
  @alexfish77923 ай бұрын

  @@mecharithm-robotics Thanks. By the way I'm actually not building robots. I'm doing game graphics programming. Requires a lot of rotation matrices and homogenous transformations. Maybe some day I'll get into robots :)

• 
  @mecharithm-robotics3 ай бұрын

  @@alexfish7792 that is a really cool application!

• 
  @alexfish77923 ай бұрын

  5:28 After a year or more of study, I finally understand what this torus is about. Thank you for the visual example.

• 
  @sabeerhussain96663 ай бұрын

  are you on linkedin?