3 ms·

From a technical side, I do not really understand why so many companies use such a complex robotic arm with 5-6 DOF. Wouldn't a SCARA in this case be much bette

From a technical side, I do not really understand why so many companies use such a complex robotic arm with 5-6 DOF. Wouldn't a SCARA in this case be much better? They're easier to program, simpler kinematics, cheaper, and more robust.

khawkins 8y ago
Actually higher degree of freedom arms with 6-7 DOF have simpler kinematics and are easier to program, in general. This is especially the case if you're using sensing and a lightly unstructured task. An end effector pose has 6 DOF (3 translation, 3 rotation). A robot with fewer than 6 DOF can't physically reach all poses in its task space, so if sensing says go here, the kinematics will often say it can get close, but not exactly. A robot with exactly 6 DOF can usually reach all poses in its task space, with a catch. Typically there are only a discrete number of joint configurations for a given pose (like 4 or 8). Thus, if you want to reach a certain pose but all joint configurations are in collision then you are in trouble. Adding just 1 more DOF, making it a 7 DOF arm, ensures that the arm can generally reach any pose with an infinite number of configurations, the last dimension used to plan avoiding collision. If you use motion planning and sensing for all of this, instead of something like a teach pendant where you carefully set all of the desired poses manually, then higher DOF is easier.
- syntaxing 8y ago
  Are high DOF robot arms easier to program? Most of the ones I've seen are usually done through human "mirroring" where the arm is moved in a generic fashion (and fine tuned by surrounding sensors) by a user and recorded similar to the pendant method you mentioned. I totally understand that a SCARA arm has a lot less flexibility. But wouldnt a low DOF arm be beneficial for an application like this, where the parts being grabbed and placed are in a highly gridded configuration?
outworlder 8y ago
Maybe the explanation is not technical. Which one looks 'better' or 'more sophisticated'? SCARA or a robotic arm? Which one is going to attract more investors?
Kagerjay 8y ago
I am not so sure either why a complex robotic arm with 5-6 DOF is necessary. It seems more suited for manufacturing in general, for handling components of varying size and complexity, or areas with limited space issues. Or the need for more modular-based processes But farms generally have lots of land / space availability. The pipeline for growing, farming, and transporting crops is fairly predictable, minus unexpected events like hurricanes/tornados/rainfall, diseases, pests, spoilage, etc With farming, everything to my understanding is mostly predictable for crop size. Farms generally specialize in one type of crop if I'm not mistaken, especially if its hydrophonics /inhouse like in the dutch farm video mentioned in this thread. SCARA makes more sense for farming

khawkins 8y ago

Actually higher degree of freedom arms with 6-7 DOF have simpler kinematics and are easier to program, in general. This is especially the case if you're using sensing and a lightly unstructured task. An end effector pose has 6 DOF (3 translation, 3 rotation). A robot with fewer than 6 DOF can't physically reach all poses in its task space, so if sensing says go here, the kinematics will often say it can get close, but not exactly. A robot with exactly 6 DOF can usually reach all poses in its task space, with a catch. Typically there are only a discrete number of joint configurations for a given pose (like 4 or 8). Thus, if you want to reach a certain pose but all joint configurations are in collision then you are in trouble. Adding just 1 more DOF, making it a 7 DOF arm, ensures that the arm can generally reach any pose with an infinite number of configurations, the last dimension used to plan avoiding collision. If you use motion planning and sensing for all of this, instead of something like a teach pendant where you carefully set all of the desired poses manually, then higher DOF is easier.

syntaxing 8y ago
Are high DOF robot arms easier to program? Most of the ones I've seen are usually done through human "mirroring" where the arm is moved in a generic fashion (and fine tuned by surrounding sensors) by a user and recorded similar to the pendant method you mentioned. I totally understand that a SCARA arm has a lot less flexibility. But wouldnt a low DOF arm be beneficial for an application like this, where the parts being grabbed and placed are in a highly gridded configuration?

outworlder 8y ago

Maybe the explanation is not technical. Which one looks 'better' or 'more sophisticated'? SCARA or a robotic arm? Which one is going to attract more investors?

Kagerjay 8y ago

I am not so sure either why a complex robotic arm with 5-6 DOF is necessary. It seems more suited for manufacturing in general, for handling components of varying size and complexity, or areas with limited space issues. Or the need for more modular-based processes But farms generally have lots of land / space availability. The pipeline for growing, farming, and transporting crops is fairly predictable, minus unexpected events like hurricanes/tornados/rainfall, diseases, pests, spoilage, etc With farming, everything to my understanding is mostly predictable for crop size. Farms generally specialize in one type of crop if I'm not mistaken, especially if its hydrophonics /inhouse like in the dutch farm video mentioned in this thread. SCARA makes more sense for farming