Verifiable Surface Disinfection Using Ultraviolet Light with a Mobile Manipulation Robot†
Round 1
Reviewer 1 Report
In this work, a human supervisor designates a surface to be disinfected. Then the robot autonomously plans a set of motions to disinfect the surface, using an ultraviolet light held in its gripper. Finally, the system displays the amount of ultraviolet radiation delivered to each part of the surface, so that the human supervisor can verify that the operation was successful. The topic is interesting and the work is solid, this manuscript is rich in content. For the benefit of the reviewer, however, a number of points need clarifying and certain statements require further justification. My detailed recommendations are as follows:
1. More explanations about the significance of motivation of this manuscript should be given especially why this method is indispensable in Section Introduction and Background.
2. Please check your reference style. And include the following related paper into discussion,
https://0-www-sciencedirect-com.brum.beds.ac.uk/science/article/abs/pii/S0140366419307960
3. In secton 4, the proposed approach should be discussed clearly with flowchart. It would be better to add one example for illustration.
4. The technical depth is not enough, and the evaluation is too simple.
The authors should enrich the two parts.
Author Response
1) We added a few sentences to this effect in the Introduction.
2) We have checked the reference style to ensure that it is approved for the journal. We have chosen not to include this citation because it is not closely related to the work reported in our paper. While both this paper and ours are about robots being used in healthcare settings, the focus of the previous paper is on the infrastructure supporting the robot, looking in particular at the wireless communications between the robot and a remote operator. Our paper is closely focused on an overall algorithmic approach for surface decontamination and on the empirical calibration of the calibrating UV-C light. While we do use wireless communications to the robot, it is not central to our approach.
3) We have added a figure to clarify the process, which appears in Figure 2.
4)We have reviewed our descriptions and believe they are sufficient for a reader to replicate the work. The system itself is relatively straightforward, and we have intentionally decided not to over-complicate or obfuscate the description. However, if there are specific parts of the description and evaluation that the reviewer would like to see expanded and can direct us to, we would be happy to do so in the final revision.
Reviewer 2 Report
1) The references list is missing.
2) compare with other work is missing.
3) Add the future work and limitations of current work
4) Abstract is not well written. The fallow is abstract is not appropriate. rewrite again.
5) Introduction part is not completed, looks very short, there is no novelty.
1) Background, Problem, and Motivation, Proposed Approach, Applications (e.g., Blockchain-Federated-Learning and Deep Learning Models for COVID-19 Detection Using CT Imaging)
6)
Author Response
1) We have verified that the reference list was included in the manuscript submitted to the journal. Other reviewers did not note this; perhaps there was some error in the paper version that was sent to Reviewer 2.
2)Section 3 (lines 80 to 127) reviews the work most closely related to that reported in the paper. While UV-C disinfection robots exist, there is relatively little published literature describing them. We believe that our related work section is complete but would be happy to expand it if the reviewer has specific references that they would like to see included.
3)We have rewritten and renamed the final section of the paper to describe our current system's limitations and better describe the future work that would make the system more generally applicable.
4)We have completely rewritten the abstract and changed the title of the article to make it clearer.
5)The introduction is, indeed, short, but we believe that it is complete. The central idea of this paper is straightforward to describe, and we made the conscious decision not to labor the point, and to keep the introduction brief and to the point.
We disagree that there is no novelty. There is no other semi-autonomous system for the UV-C disinfection of surfaces that uses a robot arm that we are aware of. We are also unaware of a similar practical analysis and modeling of UV-C light for a system such as this. If the reviewer has a specific reference to elements of this work that has been reported previously, we could be pleased to incorporate it into our paper, and to change our claims accordingly.
We have added a sentence to the introduction to state the novel contributions made by our work explicitly.
6)The intent of this comment is unclear to us. The paper referenced is unrelated to the work reported in ours, other than that it deals with COVID-19. The referenced paper deals with learning predictive models for COVID-19 infection from CT images, while ours is focused on disinfecting possibly-infected surfaces.
Reviewer 3 Report
This paper focused on the issue of using sophisticated mobile manipulation robots to perform surface disinfection in the context of the Ebola Virus Disease. It attempts to handle the robot autonomously planning a set of motions to disinfect the surface with a human supervisor. How to move from disinfecting surfaces, which are of concern for Ebola, to open space disinfection, which is important in the fight against COVID-19 have been discussed in this paper. This research work is interesting for the healthcare robot research society, and the experimental results are discussed suitably. However, this paper has several limitations and the standard is not enough, and address the following items would result in a good paper,
- The literature review is not thorough about the application and the contributions. To highlight the contributions, it suggests reorganizing the section of the related work. At least, for each contribution, it should be novel and meaningful according to a thorough literature review. In the literature analysis, it is recommended to read the following works and consider to discuss their similar application scenarios in the introduction and discussion, Improved recurrent neural network-based manipulator control with remote center of motion constraints: Experimental results; A teleoperation framework for mobile robots based on shared control.
- It is recommended to present in the first section so that it can highlight the specific scope of this article. The meaning of the assessment experiment should be highlighted.
- Figure 10 is not clear and we would suggest adding more details and labels to make it clear.
- Maybe it is better to discuss the possibility to improve the scope using deep learning to learn and control robot behavior in the introduction, for example, A Multimodal Wearable System for Continuous and Real-time Breathing Pattern Monitoring During Daily Activity; Multi-sensor Guided Hand Gestures Recognition for Teleoperated Robot using Recurrent Neural Network; Trajectory Online Adaption Based on Human Motion Prediction for Teleoperation.
- There should be a further discussion about the limitation of the current works, in particular, what could be the challenge for its related applications. To let readers better understand future work, please give specific research directions.
Author Response
1)We believe that the literature review is complete; although UV-C disinfection robots exist and are becoming widely used, there is little in the literature describing them. We have included all of the relevant references to the closely related work as far as we are aware. If the reviewer has additional specific references that we have missed, we would be happy to integrate them into our paper.
The two suggested papers TBD.
2)We have rewritten part of the introduction to make the scope and significance of the work reported more apparent.
3)Figure 10 was simply a picture of commercially available UV-C disinfection robots. Since it was not central to the paper's narrative and added little to it, we have removed this figure.
4)While deep learning is a popular control paradigm these days, we believe it is not relevant to this situation. Traditional trajectory planning techniques are sufficient for the purposes of the work described, and allow us better control over the final executed trajectory of the end-effector. While we could use deep learning approaches, we do not believe that it is necessary to do so. Further, the use of learning-based approaches makes it harder for us to guarantee coverage for a particular area, since the resulting trajectories are not as easily analyzable as those generated by a traditional coverage planner. While using established techniques, rather than the latest algorithmic advances, lessens the novelty of the work somewhat, it is also the appropriate approach for applications like the one described in this paper.
5)We have added a more explicit description of the limitations of the current work, and tried to make the proposed research directions clearer. We have renamed the last section, and rewritten parts of it to make it clearer.
Round 2
Reviewer 2 Report
Accept
Reviewer 3 Report
The authors have addressed all of my concerns. The current version can be accepted now.
