The vision system sends obj poses, but during simulation in Mech-Viz, there are planning failures and unreachable waypoints.
However, when manually moving the robot, it feels like the poses can be reached.
Robot: Fanuc five-axis robot model fanuc_m_410ib_140h
Software: Viz 1.7.2
The 5-axis robot lacks degrees of freedom, and its end effector cannot reach all possible poses in the workspace. Mech-Vision’s result is a 6-degree-of-freedom data as default, which prevents the 5-axis robot from accurately solving for the correct joint angles.
In the Mech-Vision project, the vision poses sent to Viz should be corrected by brute force, making the Z-axes of the poses perpendicular to the XY plane of the base reference frame.
Q1: Why can I manually move the robot to reach the positions?
A: The manually reached poses are just approximate poses and may not be exactly the same as the vision result poses.
Q2: 4-axis robots also have limited degrees of freedom. Why doesn’t Viz have this issue with 4-axis robots?
A: Viz has a mechanism to automatically force correct the received vision result poses when the robot is 4-axis.
Q3: Will Viz automatically correct the vision poses received by the 5-axis robot in the future?
A: It is possible, but it needs further discussion.
Q4: Since there are 4-axis robots, why are there 5-axis robots too?
A: Robots with limited degrees of freedom (3, 4, 5-axis) are often used in heavy load handling scenarios. 4-axis robots can satisfy most handling scenarios, but there are situations where a small angular deviation exists between the robot’s base and the objects, requiring more degrees of freedom. This is where 5-axis robots come into play. Another important reason is the consideration of manufacturing costs. Manufacturers often use modular designs, and 5-axis robots often reuse the wrist joint of 6-axis robots.