Digital Food Twins Combining Data Science and Food Science: System Model, Applications, and Challenges^†

Round 1

Reviewer 1 Report

The article approach a very interesting subject and is suitable for publication. 

Althouth three observations, the first one is that in line 136 the smartFoodTechnologyOWL initiative is not clearly referenced. The second one is about LRA-M loop, it´s meaning is not clear enough, at least for this evaluator.  Finally, the discussion and conclusions should go a little bit deeper on border conditions that limit the adoption of such suitable tool for food processing industries. 

I´m working with a yeast for bakeries firm, and found this methodology quite interesting, but there are plenty of limitations. You describe them but should go a  little bit further. 

Author Response

Please see the attachment, comments to reviewer 1.

Author Response File: Author Response.pdf

Reviewer 2 Report

In this paper, the author did a very systematic review on applying digital twin in food industry, and covered different aspects. This very thoughtful review provided a full picture for future food industry 4.0. However, it would benefit the audience more if the author could provided some details when describe the concepts. 

1. Digital twin has been a popular concept in recent years, and may have slightly differences in different industry. Thus,  could the author provide a definition of “digital twin” for food industry at the beginning of the manuscript?

2. Line 88: could the author provide examples of how machine learning is used in production planning

3. Line 101 and 110: “The authors of” should be “The authors”

4. Section 2 overall: could the author add an illustration figure to summary how different related work that were discussion in section applied in food industry?

5. Overall Section 3: the author covered different sub-topics such as data resource, data generation, data analysis and model selection. Could the author use subsection titles to help highlight the section focus?

6. Line 199-203: could the author provide some detailed examples of how scientific model help to forecast process steps?

7. Line206-218: could the author provide more details on process modeling? For examples, when the author discuss “describe the behavior of the bacteria”, could the author provide some details on what are the models/equations used? What parameters (cell growth rate, cell density or O2 consumption rate?) the model describes?

8. Line 239-246: in food industry,  could the author add discussions on when XAI approach (1) should be considered and when approach (2) should be considered?

9. In section 4 case study, could the author add additional cases/discussions on (1) how to validate/continue improvement of the model/digital twins? (2) data engineering examples such as data resource/data cleaning/data storage?; (3) how digital twin could help address supply chain shortage?

Author Response

Please see the attachment, comments to reviewer 2.

Author Response File: Author Response.pdf

Reviewer 3 Report

In this vision paper, the authors describe the concept of a digital food twin. Due to the variability of the raw materials, such a digital twin has to take into account not only the processing steps but also the chemical, physical, or microbiological properties that change the food independently from the processing. The authors proposed a hybrid modeling approach, which integrates the traditional approach of food process modeling and simulation of the bio-chemical and physical properties with a data-driven approach based on the application of machine learning.

This article is only a conceptual presentation of modern application of artificial neural network models, but no real experimental results were presented or any kind of a model is not developed within this paper. This makes this article a bit hard to read and understood for a broader scientific community.

The experimental results were not presented!

The model was not developed!

Thematically the work is interesting for the researchers and professionals and the proposed manuscript is relevant to the scope of the journal, but a simple example of the application of the concept should be presented in the study?

I found the article appropriate for publication in the Processes journal, but only after some modifications and clarification from the Authors.

Other than this, the overall organization and structure of the manuscript are appropriate, the paper is well written and the topic is appropriate for the journal.
The aim of the paper is well described and the discussion was well approached, its results and discussion are correlated to the cited literature data.

The literature review is comprehensive and properly done.

The novelty of the work must be more clearly demonstrated. Possible the drawback of the paper could the extent of novelty, or the main novelty in the present work, compared to the works of other researchers? In my opinion, the authors should put additional effort to demonstrate that the present work gives a substantial contribution in the research area, especially showing the comparation of the model to real experimental data.

Statistical model with the appropriate data should be presented.

The verification of the model should be performed.

Model validation is possibly the most important step in the model building sequence. 

Author Response

Please see the attachment, comments to reviewer 3.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

The Authors managed to improve the quality of the Manuscript, according to the Reviewer's comment.

I suggest the Editor to accept the Manuscript, in the presented form, for possible publication in the Processes journal.