The Application of Generative Algorithms in Human-Centered Product Development

Round 1

Reviewer 1 Report

The Authors take up the interesting new issue of efficient design of highly customized products. As it is known, the generative design uses AI-driven CAD software to generate a range of design solutions that meet a set of constraints. So the essence of the generative design method is to explore a broader range of design options generated not by human but by the computer. The authors showed designing process of "bespoke versions of a standard controller device". I was expecting to see the controller device designs that look very foreign and therefore innovative (designed by machine, not by man). Meanwhile, the external appearance of the designed controller is not surprising. The Authors show only one variant of solving the design problem. The authors also did not show how (based on what criteria) they chose the final design solution. Despite this, the article is interesting, and its strength is a comprehensive literature review.

Yours sincerely,

Author Response

Reviewer 1 notes:

• …the external appearance of the designed controller is not surprising. The Authors show only one variant of solving the design problem. The authors also did not show how (based on what criteria) they chose the final design solution.

RESPONSE: The presented solution is the result of a generative solving effort and shows segments of a basic controller configuration. To discuss the details of the design would be beyond the scope of the paper and would present issues with IP. The authors feel the paper should be more focused on the discussion of the workflow rather than detail specific design outcomes.

Reviewer 2 Report

The paper is interesting, but it needs revisions and improvements, and a major reorganization of sections and related information.

The first question to be clarified is a methodological one.

• The authors speak indistinctly of design process and development process. The authors need to make it clearer that they are only talking about the design process.
• The phases of the design process are one thing and the CAx systems used in the design process are another thing. The concept design is one thing, the tools I use in this phase are quite another.
• The described approach makes explicit and exclusive reference to the use of Grasshopper. If the authors want to make a general discourse on generative design, I recommend that they also refer to Catia, SolidWorks, Fusion360, Creo, etc. What they describe is, in fact, unique to Rhino only, while the other systems use different approaches.

Other questions:

• The state of the art regarding generative tools and their applications requires further investigation.
• Section 1.2 focuses on the topological optimization process and less on generative design process. The authors imply that generative design algorithms originate from topological optimization methods. Is this actually true? If yes, please provide bibliographic references about it. This section does not explain how the generative design process differs from the TO mentioned. What are the advantages?
• Figure 2 requires better quality as the inscriptions are not legible. Also add the source.
• is the reference to the 254 row correct?
• Figures 1, 2, 3, and 4 are not sufficiently discussed in the paper and their need to support the discussion oriented to the application of generative design tools in the field of ergonomics is not evident.
• The state of the art and methods of HCD are discussed superficially in section 2.1. Due to the central role of this topic a specific chapter should be added in section 2.
• Section 2 should be enhanced by adding more recent references about GD and GDA. Such as, for example:
  • Mountstephens, J., Teo, J. (2020). Progress and challenges in generative product design: a review of systems. Computers, 9(4), 80.
  • Jaisawal, R., Agrawal, V. (2021). Generative Design Method (GDM)–A State of Art. In IOP Conference Series: Materials Science and Engineering (Vol. 1104, No. 1, p. 012036). IOP Publishing.
  • Barbieri, L., Muzzupappa, M. (2022). Performance-Driven Engineering Design Approaches Based on Generative Design and Topology Optimization Tools: A Comparative Study. Applied Sciences, 12(4), 2106.
  • De Crescenzio, F., Fantini, M., Asllani, E. (2022). Generative design of 3D printed hands-free door handles for reduction of contagion risk in public buildings. International Journal on Interactive Design and Manufacturing (IJIDeM), 1-9.
• In section 3 the authors describe the key steps that were followed in the construction of a VR controller architecture for the PRIME-VR2 project. These descriptions focus on the results obtained without providing the sufficient details necessary for the reader to understand and replicate the procedure. In fact, image 5 (left) is only an exemplification of the arc diagram and therefore does not add useful information to support the case study. It would be necessary instead to show as it has been realized and which is the arc diagram realized specifically for such case study.
• Figure 7 (left) is unclear and unreadable.
• For better understanding and communication of the research work, it would be appropriate to move Section 4, which describes the proposed method, before the case study.
• Most references lack basic information. For example, it is not possible to infer whether they are journal publications, conference proceedings, or books.

Author Response

Reviewer 2 notes:

• The authors speak indistinctly of design process and development process. The authors need to make it clearer that they are only talking about the design process.

RESPONSE: For clarity, we now refer to the “(product) design development process” as we consider a generative design process as also part of design development work

• The described approach makes explicit and exclusive reference to the use of Grasshopper. If the authors want to make a general discourse on generative design, I recommend that they also refer to Catia, SolidWorks, Fusion360, Creo, etc. What they describe is, in fact, unique to Rhino only, while the other systems use different approaches.

RESPONSE: These tools are not unique to Rhino as there are other CAD platforms that facilitate similar workflows. The difference with other CAD packages has now been included within the discussion for clarity.

• The state of the art regarding generative tools and their applications requires further investigation.
• Section 1.2 focuses on the topological optimization process and less on generative design process. The authors imply that generative design algorithms originate from topological optimization methods. Is this actually true? If yes, please provide bibliographic references about it. This section does not explain how the generative design process differs from the TO mentioned. What are the advantages?

RESPONSE: This has been clarified. While TO and generative design are distinct methods, the modelling and solution finding strategies follow similar logic and there is overlap between the two. Furthermore, by extending this discussion we feel this addresses the concern that generative design has not been adequately analysed.

• Figure 2 requires better quality as the inscriptions are not legible. Also add the source.

Image removed

• is the reference to the 254 row correct?

Removed for clarity

• Figures 1, 2, 3, and 4 are not sufficiently discussed in the paper and their need to support the discussion oriented to the application of generative design tools in the field of ergonomics is not evident.

RESPONSE: Figures 2 and 3 have been removed to make the discussion more focused. We feel figures 1 and 4 should remain to provide vital context to the paper and the development and thinking behind the work. Extra detail has been added to supplement their inclusion

• The state of the art and methods of HCD are discussed superficially in section 2.1. Due to the central role of this topic a specific chapter should be added in section 2.

RESPONSE: Material has been added in sections 1.1 and 2.1 to focus on HCD methods and highlight their relevance to generative design thinking

• Section 2 should be enhanced by adding more recent references about GD and GDA. Such as, for example:
• Mountstephens, J., Teo, J. (2020). Progress and challenges in generative product design: a review of systems. Computers, 9(4), 80.
• Jaisawal, R., Agrawal, V. (2021). Generative Design Method (GDM)–A State of Art. In IOP Conference Series: Materials Science and Engineering (Vol. 1104, No. 1, p. 012036). IOP Publishing.
• Barbieri, L., Muzzupappa, M. (2022). Performance-Driven Engineering Design Approaches Based on Generative Design and Topology Optimization Tools: A Comparative Study. Applied Sciences, 12(4), 2106.
• De Crescenzio, F., Fantini, M., Asllani, E. (2022). Generative design of 3D printed hands-free door handles for reduction of contagion risk in public buildings. International Journal on Interactive Design and Manufacturing (IJIDeM), 1-9.

These references with discussion have been added

• In section 3 the authors describe the key steps that were followed in the construction of a VR controller architecture for the PRIME-VR2 project. These descriptions focus on the results obtained without providing the sufficient details necessary for the reader to understand and replicate the procedure. In fact, image 5 (left) is only an exemplification of the arc diagram and therefore does not add useful information to support the case study. It would be necessary instead to show as it has been realized and which is the arc diagram realized specifically for such case study.

RESPONSE: The work is protected by intellectual property, so the descriptions provided must remain high-level/non-specific. Additionally, the aim of this paper is to present a more general/theoretical approach as opposed to something that can be precisely replicated.

• Figure 7 (left) is unclear and unreadable.

RESPONSE: The image is used to provide a general view of how the different information can connect in a structurally complex way. It is not intended to provide specific information

• For better understanding and communication of the research work, it would be appropriate to move Section 4, which describes the proposed method, before the case study.

RESPONSE: The section has been swapped with section 3 which should provide an improved communication or narrative.

• Most references lack basic information. For example, it is not possible to infer whether they are journal publications, conference proceedings, or books.

References have been amended to provide more complete information

Reviewer 3 Report

The manuscript proposes an interesting study on ergonomics design by means of new technologies such as generative algorithms, rapid prototyping, and additive manufacturing technologies.

Despite such a goal that fits very well with the journal topics, the manuscripts presents several critical flows that need to be addressed.

Firstly, in the introduction, research motivations need to be discussed more in detail, demonstrating how the current study can augment knowledge in the field of human-centred design. In such a context, the definitions and current achievements from different perspectives on human-centred design should be taken into account (e.g. see http://www.ieomsociety.org/singapore2021/papers/1307.pdf; https://doi.org/10.1080/09544828.2016.1235262; https://doi.org/10.1080/00140130400029167).

Then, section 1 should be split into different sections, moving subsections 1.1 and 1.2 in a background analysis section.

Then, when discussing the research methodology a description of the proposed research approach is needed.

The discussion of results also needs to be augmented, bringing to light the methodological and practical implications of the study.

Author Response

• Firstly, in the introduction, research motivations need to be discussed more in detail, demonstrating how the current study can augment knowledge in the field of human-centred design. In such a context, the definitions and current achievements from different perspectives on human-centred design should be taken into account (e.g. see http://www.ieomsociety.org/singapore2021/papers/1307.pdf; https://doi.org/10.1080/09544828.2016.1235262; https://doi.org/10.1080/00140130400029167).

RESPONSE: These works have been noted with interest and discussion of them has now been included within sections 2 and 3

• Then, section 1 should be split into different sections, moving subsections 1.1 and 1.2 in a background analysis section.

RESPONSE: Sections 1.1 and 1.2 have now been combined into one section of general background, which should improve the narrative

• Then, when discussing the research methodology a description of the proposed research approach is needed.

RESPONSE: Agreed. A statement addressing the research approach and methodology has been added within section 3 and the introduction

• The discussion of results also needs to be augmented, bringing to light the methodological and practical implications of the study.

RESPONSE: More discussion of the practical and theoretical implications have been added with a new section 4.3 within the amended chapter 4

Round 2

Reviewer 2 Report

No futher suggestions

Reviewer 3 Report

The Authors have improved the quality of the manuscript satisfactorily. Hence, it can be considered for publication.