Moving toward Smart Manufacturing with an Autonomous Pallet Racking Inspection System Based on MobileNetV2

Round 1

Reviewer 1 Report

- Manuscript doest not have keywords

- Images are not good quality

- Must improve english, consider proof reading

- Tables misses fomatting according template

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

1.      Introduction should be enriched with more details about Industry 4.0, Smart Manufacturing and Digital Supply Chain.

Please check the following publications in the abovementioned fields in order to present the manufacturing situation, to focus on the significance of safety in shop floor level, present the research question and describe the proposed solution:

·        Mourtzis, Dimitris (Book). Design and operation of production networks for mass personalization in the era of cloud technology (2021), pp. 1 - 393. DOI: 10.1016/C2019-0-05325-3

·        Deja, M., Siemiątkowski, M.S., Vosniakos, G.C. and Maltezos, G., 2020. Opportunities and challenges for exploiting drones in agile manufacturing systems. Procedia Manufacturing, 51, pp.527-534.

2.      Figure 2 should be presented in a more scientific manner including the information as well as the material flow. It is suggested to add the necessary ICT infrastructure that will be used in this framework.

3.      It is recommended to begin a Section with the paragraph followed by the Figure. Thus, please elaborate Section 2.6.

4.      In Figure 7, please add as previous more technical information such as communication protocols, network details etc.

5.      Figure 8 does not present the proposed System Architecture. It is probably a typo from the previous Figure Section. Please elaborate.

6.      The authors state that the paper led to the proposal of the first Deep Learning based real-time damage detection of pallet racking. Please indicate in a more detailed manner the real-time term. Which is the latency of the used network?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

The research given in the manuscript is significant for logistics in-house processes without any doubt. Nevertheless, the manuscript should be developed before publication.

In fact, the literature review in the manuscript is rather poor. Please, enrich your contribution with a proper literature review. The review should be summed up with a research gap/-s and then the contribution.

Whereas reviews related to such topics of sensors and racking, some other researchers may be listed as well. Securing safety by monitoring of technical parameters in warehouse racks in warehouses should be searched. The processes (the ones using AR technologies with high potential of neural networks / deep learning applied) may interest the Authors (especially since the processes of production and logistics, e.g. rack storing are relatively similar). Please, check for example: Szajna, A.; Kostrzewski, M.; Ciebiera, K.; Stryjski, R.; Woźniak, W. Application of the Deep CNN-Based Method in Industrial System for Wire Marking Identification. Energies 2021, 14, 3659. https://doi.org/10.3390/en14123659; Neupane, D.; Kim, Y.; Seok, J.; Hong, J. CNN-Based Fault Detection for Smart Manufacturing . Appl. Sci. 2021, 11, 11732. https://doi.org/10.3390/app112411732; Szajna, A.; Stryjski, R.; Woźniak, W.; Chamier-Gliszczyński, N.; Kostrzewski, M. Assessment of Augmented Reality in Manual Wiring Production Process with Use of Mobile AR Glasses. Sensors 2020, 20 (17), 4755, pp. 1-26. https://doi.org/10.3390/s20174755

The mentioned authors' experience may become strong support in the research of the manuscript's Authors.

The structure of the manuscript should be developed according to guidelines. For example, the methodology, and the results should be presented in a separate section.

The conclusion should be developed. Research limitations and future research ideas should be given.

The abbreviation should be defined with their first use.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

The submitted article proposes autonomous racking inspection mechanism using a MobileNetV2-SSD architecture.

A solution is proposed that is affixed to the adjustable cage of a forklift truck, enabling adequate coverage of racking in the immediate vicinity.

Proposed approach leads to a classifier that is optimized for deployment onto edge devices, providing real-time alerts of damage to forklift drivers, with a Mean Average Precision of 92.7%.

I consider submitted article partly be original.

It contains an adequate amount of literature used.

The current scientific literature is used in the article.

The conclusion section is quite short. This section does not reflect the entire content of the manuscript, including the analysis of the results obtained. This section should be extended.

In the present article would be appropriate to clarify how experimental identifications are used in practice.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The authors have revised the manuscript taking into consideration the comments raised by the reviewers. No further comments are raised.

Reviewer 3 Report

Almost all of my remarks were considered, therefore the Editors may undertake their final decission.