Implementation and Evaluation of Dynamic Task Allocation for Human–Robot Collaboration in Assembly

Round 1

Reviewer 1 Report

I congratulate the authors on a very extensive and well-organized research and its presentation.

First of all, the topic is very interesting and from the perspective of practical application it has great value, because the technology of collaborative robots has yet to enter the industrial space on a large scale. Collaboration is, generally speaking, a very complex phenomenon, viewed at the systemic level. When talking about collaboration between biological and engineered systems, in this case robots, it becomes even more complex for understanding, modeling and also for planning and control.

The proposed approach of Dynamic Task Allocation in domain of Collaborative Robots can be considered methodologically complete and obviously practically applicable, but there are some minor weaknesses that need to be corrected. In this regard, I have two proposals, that is, requests to the authors:

1. The essence of the approach is explained in Figure 7. Flowchart of the proposed methodology for dynamic task allocation. A very messy information flaw, not well structured, with many branches and feedback loops, which make it difficult to follow logically and consequently reduce tractability. It is necessary to supplement this flowchart with a new flowchart. That would be a BASIC Flowchart, containing only basic logic/process blocks and focusing on very basic information flows. The flowchart shown in Figure 7 should be retained, but that flowchart would serve as a more detailed explanation of the methodology provided by the BASIC flowchart. This is just a suggestion. I leave it to the authors to decide how to improve the comprehensibility of the diagram explaining the proposed approach.

2. Focus is on DYNAMIC Task Allocation. Dynamical systems are time systems and they are, in a mathematical but also essential sense, represented by a state-transition function (in this particular case it is the state-space of the collaborative (dynamical) system consisting of humans and robots) and an output function. The behavior of dynamic systems represented by a state-transition function is not intuitive (especially in the case of complex nonlinear and discrete systems with a large number of internal states and a large number of inputs and outputs) and this is the reason why the analysis of dynamic system is most often based on modeling and simulation. In the proposed method there are many intuitive elements in decision-making sequences. Please comment on this issue and, in connection with it, the issue of observability, controllability and stability of the algorithm on which the proposed Dynamic Task Allocation approach is based.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Overall, the manuscript is well written. The current manuscript is interesting and novel, so due to that the manuscript could of valuable contribution to the scientific community. Generally, the study appears to be sound, well-designed and could be also of practical value. Therefore, the undertaken issue is interesting and worth of disseminating. However, I suggest the authors add a few additions and corrections to this paper:

The introduction is well written with a large number of citations. However, it seems extremely long with a large amount of general information that does not contribute significantly to this manuscript and hypothesis. Thus, I would advise to reduce the introduction by 10-20% to an acceptable size.

It may be worth describing the areas of production where HRCs are applicable so that readers better understand the importance of the manuscript.

Most of the figures in the manuscript are difficult to read due to the small print. Please try zooming in on the illustrations and explanatory text. This is especially true for figures 8, 9, 12, 15 and 19.

Very little information is provided on the repeatability of the experiment, it is worth expanding this section.

Author Response

Please see the attachment.

Author Response File: Author Response.docx