Optimizing Kinematic Modeling and Self-Collision Detection of a Mobile Manipulator Robot by Considering the Actual Physical Structure

Round 1

Reviewer 1 Report

In this paper, a modified kinematic modeling method for mobile manipulators is proposed. The objective robot is a serial configuration with motors at the joints, and the distal link shifting on the joint axis. A relatively low computational cost self-collision detection method that considered the shape of the links and motors is proposed. The superiority of the proposed method is demonstrated by comparing its computation time and self-collision detection accuracy with those of related methods and typical methods.

The problem addressed in this paper regards the real-time motion planning of mobile manipulators. A more accurate self-collision detection and a lighter computational cost will allow the use of a larger workspace. The paper is well organized. The methods and the flow of calculations are explained very carefully. The content of the paper can be acceptable. However, the presentation can be improved to make the proposal and results easier to understand.

• In the introduction, the author explains the problem of the manipulator kinematics using a specific structure of the robot without a definition of the robot nor illustration. For example, "joints 2 and 3" in Line:73 can only be understood by readers who know the specific structure of the robot. I recommend explaining it by citing Figs. 1 or 2.
• Line538: What is "step2" refer to? Description of “Step2” in section 3.3.1 seems not to match the corresponding. it should be avoided using the same term, as it confuses.
• - Line 594: "Formula 12" seems to be a mistake. I think it means (18) or (19).
• i and j in equations (18) to (21) make it easy to mislead that represent "non-adjacent elements". I recommend defining the range of the parameters and the relationship of i and j as the equations.
• For the evaluation of the accuracy of the approximation in section 5.3.1, I expected a quantitative comparison with previous studies [36], but only qualitative by comparing the figures were provided. The quantitative comparison using such as volume of dead space would be useful for future studies.
• Units should be indicated in all of Tables 1 to 4.
• Line788: Maybe not T^2_d, but T^1_d.
• Line816: NS_i should be SN_i, and NC_i should be CN_i.
• Lines857-860: The explanation of the spherical objects' color may be about Fig. 8.
• Section 5.4 provides the result and discussion in detail. But it is not easy to imagine the situation especially about of discussion 4) (Line:959). For example, simply indicating the self-collision detection points in Fig. 9(a,c), Fig. 10(a,c), and Fig. 11(a,c) would help the reader to understand.

Author Response

 "Please see the attachment."

Author Response File: Author Response.pdf

Reviewer 2 Report

Please correct the following issues:

1. In figures 1, 2 and 3, the coordinate systems has barely visible notations. Please update the images.

2. In text, at line 221, there is a reference to “other studies”. Please add a reference for these.

 

Comment and advise on future work:

Regarding the approximation of paragraph starting at line 436, you can use the finite element method to break down bigger components into finite sub-components or basic elements. Your strategy to use multiple sphere elements has the correct direction, but you should investigate using cylinders, boxes or other geometric shapes, for big elements, otherwise you get a bumpy surface, as seen in figure 7, with multiple conditions to check, for each sphere. Spheres should be used on corners or on other complex sections, to radially increase the collision avoidance space. All of these can then be used to compute a mesh of 3D forbidden space. As a conclusion, you need to consider other shapes as seen in https://journals.sagepub.com/doi/pdf/10.1177/1729881418787075, plus a strategy to early detect near collision is needed to avoid checking all conditions on each step.

 

Author Response

 "Please see the attachment."

Author Response File: Author Response.pdf