Відео

The Maslow is also way more accurate.

Entirely apples to oranges comparison. Maslow not only uses four actuators, but it's also using them within a different reference; Maslow essentially pulls itself to a coordinate because all of the actuators are clustered on the sled and pull towards an anchor. Whereas this has external actuators that pull an end effector to a coordinate via pulleys that don't exactly act as anchors, given that the angle between a given pulley and the end effector sled slips around the pulley's surface, the 'anchor' essentially rotates around a pivot and the 'anchor' is the tangential point of the pulley's surface. Give it some develop to refine the coordinate system in relation to motor spindle rotation, and it'll become just as accurate; also scale it, this is essentially yarn on pegboard, plus there's a decent amount of slop in the yarn, a better material that can more easily be pulled taut will provide better results. Theoretically the Maslow system could be simplified to two actuators, just use a through-spindle or double-spindle setup on two motors, though it's more difficult to account for grid skew (notice how these cable sleds always have for independent angles between anchor and sled) this way, but there's also the issue that you need to design around the tool, which is why the Maslow system uses four independent actuators as then you don't have to deal with routing things around the tool itself. What the demonstration here shows is encapsulating the four-cable sled within a system that uses one less actuator by manipulating pulleys to achieve similar results; you can also theoretically reduce this system to two actuators, but the third actuator here is more for keeping tension on the cable than anything, basic intuition will tell you that when you look at the routing and see that tackles are being used as a tensioner, this more or less makes this system a 2+1 actuator system considering the extra actuator only exists for tensioning rather than manipulation. Notice how tension is important, reducing Maslow to two actuators will also cause tensioning issues, having two actuators tensioning two ends (inversely in this case) works much better than constantly adjusting tension to be 'just right' on a through-spindle or double-spindle, this is akin to the issue that belt-driven CNC platforms have where you need to adjust tension to be 'just right,' the only real workaround to this would possibly be synchromesh cable but I also don't know the dimensional qualities of synchromesh, I'm not sure if it has any elasticity or if it likes to droop; and beyond this the only real solution would be a dynamically adjusting tensioning system, hence the third actuator in this demonstration tensioning both of the manipulating actuators' cables, hypothetically the Maslow two-actuator system could be tensioned by a cam-pulley or and spring, potentially tying the two opposite corners together and using an active tensioner to create a tensioned loop, but this just adds complexity especially when Maslow is made to be simple. Again, very different systems, apples to oranges is being generous, it'd be better to refer to this as 'comparaison pomme-pomme de terre,' if you understand French, essentially an apple to potato comparison, given how vastly different the two systems are, literally the only commonality is four cables between the anchor and sled, otherwise every other aspect is entirely different.

@@xaytana You lost me at "synchromesh".

@@xaytana PS - for ANY two-DoF system, having more than two positioners shall over-determine the system. If you are using an extra actuator as a variable tensioner, then fine, but it does not matter if you have one or two or four extra. Same difference. Two positioners shall define the DoF exactly, the rest shall merely keep the strings taut. The onus to explain the novelty and innovativeness is upon the authors of the publication; it is not for me to guess what may or may not have been "magicked" into the pulleys (and hence my original comment).

@@AdityaMehendale Can you tell me what articles about plane four cables two degrees of freedom

A custom made controller is implemented on Matlab Simulink Realtime to controle each axis through its Yaskawa Sigma-7 Drive

This robot is closer to the Hexa kinematics with the difference that it has coincidental joints on the mobile platform. With assumed planar platform, the number of exact complex solutions could be 36 for the forward kinematics problem.

The motors are EC 90 flat 260 W from Maxon

@@robotics_lab_ulavalthanks a lot

Shock absorbers, according to the video :)