This is the only video I've found so far which actually explains singularities. Rather than just saying "they're bad" or jumping straight into the mathematical formulae.
@Armin Azami thanks so much! In January we plotted out the physical locations of singularities for common 3-DIF robots. demonstrations.wolfram.com/RobotSingularitiesInThreeLinkManipulators/ Do you think a video of this would be helpful?
Finding out that they use the wrist singularity at trade shows because it looks cool makes it, like, 100x clearer to me why I'm having such a hard time understanding why singularities are problematic. I'm sitting here thinking "it looks like it's moving normally to me."
It was delightful to make this. I have a new demo at demonstrations.wolfram.com/RobotSingularitiesInThreeLinkManipulators/ Would it be helpful to make a video of this?
I work in industry and, at least according to the manuals, ABB robots do not have singularity issues, programming wise. Mechanically, yes, those singularities exist, I'm not saying that. From what I have read, I may be wrong, is that this is because the ABB robots use quaternion mathematics for the pose instead of a homogenous transform matrix.
Thanks @capnthepeafarmer All manipulators (even ABB ones) have singularities on the boundary of their workspace. In general, the singularities are not due to the representation (quaternions or rotation matrices, etc), but rather relate to what velocities can be generated at a given pose of the robot. As a side question -- how many DOF do your ABB robots have? Redundant manipulators can avoid some singularities in much the same way you can avoid gimbal lock by adding another gimbal and using a motor to ensure the new gimbal is not aligned with the others en.wikipedia.org/wiki/Gimbal_lock. With that said -- can you share a link to the manual page that makes this claim? Maybe they include an explanation of what they mean. My student and I made an online demo to show where common singularities are in 3-axis robots: demonstrations.wolfram.com/RobotSingularitiesInThreeLinkManipulators/
@@AaronBecker Unfortunately it's not a public document, but I think you can find some older versions on the web. It's ABB document 3HAC050947-001 "Rapid Overview" section 2.7 "Singularities". Here is an excerpt for "Program execution through singularities": "During joint interpolation, problems do not occur when the robot passes singular points." "When executing a linear or circular path close to a singularity, the velocities in some joints (1 and 6/4 and 6) may be very high. In order not to exceed the maximum joint velocities, the linear path velocity is reduced." Here is a link to the instruction set for ABB robots. Section 3.52 shows how it uses the orient and pose in quaternions. tinyurl.com/pn3ph22w It doesn't specifically state it here that is is because of quaternions. I think that might have been in a white paper. I will try and find that. The white paper was mainly trying to establish that the use of quaternions was less computationally intense than HTM.
@@capnthepeafarmer Thanks -- the manual is fun to read 🤖! Page 693 is important because it explains how they handle singularities -- they either divert the robot position away from them, divert the wrist orientation away from them, slightly divert the robot while limiting the maximum tool velocity, or stop the robot before it enters a singular area. 👍for Quaternions. Introduction 2 robotics already covers a lot of ground, so I don't introduce quaternions, but they have many benefits as a representation. They won't eliminate a singularity, but they do improve numerical stability and require less memory.
7:02 why would you want to operate or do anything close to that point ? If you limit for example blue lever to +/- 45 degrees you do not have singularity and you have excellent performance ??
@Milorad you've suggested a valid solution -- but you've limited your range of motion! What if the item your robot needs to manipulate is located beyond the 45 degree mark? Engineers are always dealing with tradeoffs, and this is one of those times.
Because the entry for yaw in our Jacobian is 1, yaw is always controllable (it is never in a singularity). That makes it uninteresting in a lecture about robot singularities.
![GUILTY GEAR -STRIVE- Season Pass 4 Playable Character #1 [Queen Dizzy] Trailer](http://i.ytimg.com/vi/nXDsAgYXXv4/mqdefault.jpg)
This is the only video I've found so far which actually explains singularities. Rather than just saying "they're bad" or jumping straight into the mathematical formulae.
Best explanation of Singularity.
For real. With to-the-point calculus to make sense of it. Top notch !
Finally, I found someone understandably and practically teaches Singularity.
@Armin Azami thanks so much! In January we plotted out the physical locations of singularities for common 3-DIF robots. demonstrations.wolfram.com/RobotSingularitiesInThreeLinkManipulators/ Do you think a video of this would be helpful?
Dream of my life would be to have the basics of math so I could attend a class of that and follow it. Beautiful stuff.
Finding out that they use the wrist singularity at trade shows because it looks cool makes it, like, 100x clearer to me why I'm having such a hard time understanding why singularities are problematic. I'm sitting here thinking "it looks like it's moving normally to me."
this is so cool, thank you so much for sharing!
It was delightful to make this. I have a new demo at demonstrations.wolfram.com/RobotSingularitiesInThreeLinkManipulators/ Would it be helpful to make a video of this?
@@AaronBecker Yes exactly! Thank you for the sharing.
Where are lecture 12, 13 or the other missing lectures? Great series by the way!
The full playlist is here ruclips.net/video/NUJl9Ju7qNc/видео.html You can find lecture 12 at ruclips.net/video/RHm7XALM0Fc/видео.html
@@AaronBecker Thank you. Looking forward to binging this playlist xD
i am embaressed now that i justed wanted the simple explanation i am sry..... but i got it, so thanks!
I work in industry and, at least according to the manuals, ABB robots do not have singularity issues, programming wise. Mechanically, yes, those singularities exist, I'm not saying that. From what I have read, I may be wrong, is that this is because the ABB robots use quaternion mathematics for the pose instead of a homogenous transform matrix.
Thanks @capnthepeafarmer All manipulators (even ABB ones) have singularities on the boundary of their workspace. In general, the singularities are not due to the representation (quaternions or rotation matrices, etc), but rather relate to what velocities can be generated at a given pose of the robot. As a side question -- how many DOF do your ABB robots have? Redundant manipulators can avoid some singularities in much the same way you can avoid gimbal lock by adding another gimbal and using a motor to ensure the new gimbal is not aligned with the others en.wikipedia.org/wiki/Gimbal_lock. With that said -- can you share a link to the manual page that makes this claim? Maybe they include an explanation of what they mean. My student and I made an online demo to show where common singularities are in 3-axis robots: demonstrations.wolfram.com/RobotSingularitiesInThreeLinkManipulators/
@@AaronBecker Unfortunately it's not a public document, but I think you can find some older versions on the web. It's ABB document 3HAC050947-001 "Rapid Overview" section 2.7 "Singularities". Here is an excerpt for "Program execution through singularities": "During joint interpolation, problems do not occur when the robot passes singular
points." "When executing a linear or circular path close to a singularity, the velocities in
some joints (1 and 6/4 and 6) may be very high. In order not to exceed the maximum
joint velocities, the linear path velocity is reduced."
Here is a link to the instruction set for ABB robots. Section 3.52 shows how it uses the orient and pose in quaternions. tinyurl.com/pn3ph22w
It doesn't specifically state it here that is is because of quaternions. I think that might have been in a white paper. I will try and find that. The white paper was mainly trying to establish that the use of quaternions was less computationally intense than HTM.
@@capnthepeafarmer Thanks -- the manual is fun to read 🤖! Page 693 is important because it explains how they handle singularities -- they either divert the robot position away from them, divert the wrist orientation away from them, slightly divert the robot while limiting the maximum tool velocity, or stop the robot before it enters a singular area. 👍for Quaternions. Introduction 2 robotics already covers a lot of ground, so I don't introduce quaternions, but they have many benefits as a representation. They won't eliminate a singularity, but they do improve numerical stability and require less memory.
7:02 why would you want to operate or do anything close to that point ? If you limit for example blue lever to +/- 45 degrees you do not have singularity and you have excellent performance ??
@Milorad you've suggested a valid solution -- but you've limited your range of motion! What if the item your robot needs to manipulate is located beyond the 45 degree mark? Engineers are always dealing with tradeoffs, and this is one of those times.
@@AaronBecker longer arm.... there are more solutions, and I do understand it is nice to know the problems and where they are in order to avoid them.
Why we are not interesting yaw?
At what time is yaw mentioned?
@@AaronBecker 14:15 thank you
Because the entry for yaw in our Jacobian is 1, yaw is always controllable (it is never in a singularity). That makes it uninteresting in a lecture about robot singularities.
@@AaronBecker Thank you so much now clear :))