Trajectory Planning of Dual-Robot Cooperative Assembly
Round 1
Reviewer 1 Report
- I miss a more detailed description of the SW and HW equipment used in the experiment.
- ABB uses its proprietary ABB RobotStudio software for trajectory planning which is free as ROS. Does this original software ABB RobotStudio not contain similar features ? If so what are the differences between the results ?
- When the authors were planning the trajectories, did they consider other requirements such as minimum energy consumption, degradation, repeatability ?
- Don't the authors see a limitation to their approach in using the ROS OS for compatibility in potential industrial use ?
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The article is concerned with the development of a method for planning the trajectories of two industrial robots jointly performing assembly operations in a working environment with obstacles. The advantage of the method is the ability to automate the programming of assembly operations performed by two robots, as well as high computational speed and high probability of finding a path.
At the same time, the following critical comments can be made.
1. The novelty of the proposed RRT-Connect method and its difference from the original method are not shown clearly. For a better understanding, this algorithm should be given in the text of the article. It is also unclear exactly how this algorithm ensures the collision avoidance for two manipulators working in the same work area.
2. Algorithm 1. It is not clear how the index r is changed during the operation of the algorithm.
3. It should be explained how the absence of collisions is guaranteed after smoothing the trajectories, since the resulting trajectories change. At the same time, it is not clear why some parts of the trajectories are changed after smoothing, and some parts are the same (Fig.9), although smoothing should be performed for the full trajectory.
4. Calculation of acceleration. Accelerations are considered as the ratio of an increment of the average velocities of the coordinate change to the sum of the time intervals of movement over two sections (eq.8). That is, it is assumed that the speed should change uniformly from the beginning of the movement from the point iqj-1 to point iqj+1. However, in the process of movement, when manipulator moves from one segment to another, acceleration will act for a short period of time. This will lead to the fact that the real acceleration will be much greater. Therefore, the proposed method of estimating acceleration is questionable.
5. The experiment was performed in an environment without obstacles, which does not allow to fully confirm the effectiveness of the method, the feature of which is the planning of trajectories in a complex working environment with obstacles.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
I want to congratulate the authors on a high-quality paper.
The paper is well organized but the English can be improved.
The overall quality of the paper is high and provides more than enough information to understand the experiments and how the authors achieved their results.
My observations are primarily on English errors and some suggestions about the used words. These are:
- Line 61 typo: "dose" instead of "does".
- Line 128, please check and correct: "There is no need to extend the object"
- Please correct English at line 137.
- Please correct the English at line 190.
- Please rewrite the phrase in lines 299-300. As a suggestion, "median" should be replaced with "average" while "the least" should be replaced with "the lowest".
- Line 322, the title should be corrected.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
