Transferring Fire Modelling Sciences into Augmented Reality: A Realistic and Safe Reconstructed Fire Scenario

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The article is interesting and well-prepared. Nevertheless I have a few comments:

1. In chapter 2. Metodology authors recall Lacrose building in Melbourne fire as well-documented event, therefore in my opinion some articles about this subject should be cited. What is more authors indicate that they reconstructed the event, but in the further part of the manuscript I have no find comparision of the real situation with the simulation.
2. In the presented research authors combines many computer programs (PyroSIm, CATIA, Blender, Paraview, etc.), but not everyone must be familiar with all of them, so in my opinion even website adresses of these programs should be given.
3. It is not clear what means some symbols in Fig. 2. Workflow for converting.... (.q, .xyz, .x3d, .fbx)
4. The figures numeration is mistaken (two figures have number 1), caption of figures should  be placed directly below the figure, not at the next page.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript presents an innovative approach to integrating fire dynamics simulations with augmented reality (AR) to enhance fire safety training and emergency preparedness. The study is well-structured and demonstrates significant potential for advancing research in fire engineering and human behavior analysis during evacuations. The combination of Fire Dynamics Simulator (FDS) with AR visualization is novel, and the user perception study adds an empirical dimension to the work. However, the manuscript requires several improvements before it can be considered for publication in this journal.

The manuscript lacks a deeper discussion on the limitations of current fire training methods (e.g., VR vs. AR, traditional drills vs. simulations). The authors should elaborate on existing research gaps and how their work addresses them.

References on recent advancements in AR applications for emergency training and human behavior modeling could be expanded.

One of the major issues when talking about another method enhance fire safety training and emergency preparedness is how psychological factors are influencing decision-making in AR environments. How does immersion in AR compare with traditional training in terms of cognitive impact and risk perception? A paragraph here will enhance this understanding.

The authors state that FDS is used for fire dynamics simulation, but they do not provide sufficient details about the numerical setup. Key missing aspects:

• Grid independence study (how was mesh resolution selected and validated?), was a D* calculated?
• Turbulence modeling approach (which sub-models were used, e.g., Large Eddy Simulation or Very Large Eddy Simulation? Or?)
• Combustion and smoke modeling (what assumptions were made about fire growth and material properties?) like number of layers, thickness, pyrolysis properties? Reactions / byproducts? What reaction was chosen from the library? Or the authors created one reaction for the current simulation?

The workflow for transferring simulation results into AR via Paraview, Blender, and Unity3D is described, but it lacks technical validation. How does the transformation from FDS data to AR affect accuracy?

The authors mention using Vuforia for tracking, but how does the accuracy of AR object placement impact user perception of fire severity and escape routes.

The manuscript does not include a proper validation of the fire simulations. The authors should compare their results with experimental data or prior numerical studies that were validated.

The discussion on model accuracy should include potential sources of error (e.g., assumptions in fire growth rate, AR rendering limitations).

Were the participants from a relevant demographic (e.g., fire safety professionals, students, general public)?

Was the sample size of the participants statistically justified?

Did the survey account for bias (e.g., familiarity with AR or fire safety training background)?

The authors claim that "77% of participants preferred AR," but without confidence intervals or significance tests, it is difficult to determine whether this preference is statistically meaningful.

In Figure 1, is it not clear why so much of the room’s surface is filled with blocks? Are these block rooms or cubicle? Does the smoke enter inside or not? Are people living inside?  

Figure captions should be more descriptive, indicating what specific insights can be drawn from each figure.

The authors should explicitly discuss limitations of their approach and future work.

.

 

Author Response

Please see the updated attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have added a discussion contrasting AR with VR and traditional fire drills. However, the comparison remains somewhat superficial. A more detailed discussion on the cognitive and psychological impacts of AR immersion versus conventional training methods would further strengthen this section.

The authors have touched upon risk perception and situational awareness but have not provided a thorough comparison of cognitive effects between AR and traditional methods.

The authors have confirmed the use of Large Eddy Simulation (LES), but it is expected that they explicitly specify the exact turbulence model applied within FDS. Since FDS includes multiple formulations, such as Standard LES, Very Large Eddy Simulation (VLES), Simple VLES, and Direct Numerical Simulation (DNS), it is important to clarify which model was selected.

The authors have described the workflow more clearly, but they have not quantitatively assessed how much accuracy is lost when translating FDS data into AR visualizations.

You could include a graph of smoke opacity vs. distance to clearly illustrate how visibility decreases depending on the scenario. This would enhance the understanding of how smoke impacts evacuees' perception and decision-making during fire emergencies.

In the Author Contributions section, it is stated that “Jason C.S. Wong” was responsible for Validation, but this aspect is not explicitly detailed in the manuscript.

If no validation has been performed in a clear and structured manner, how can we trust the accuracy of this numerical simulation and rely on the conclusions presented in the paper? It is important to demonstrate that the fire dynamics simulation is realistic and scientifically accurate, either by comparing it with experimental data, previous validated studies, or providing a quantitative error analysis. Without such validation, the credibility of the findings remains uncertain.

In the sentence:
"Similarly, for Case 2, fine mesh grids are generated near the fire source with dimensions of 0.05 m × 0.05 m × 0.04 m, while the largest grid sizes are 0.4 m × 0.4 m × 0.3 m."

There is an inconsistency in terminology. The term "mesh grid" is not standard in the context of FDS. Since FDS uses a structured Cartesian mesh, the correct term should refer to cells or mesh resolution. The authors should clarify whether they mean individual cells or regions of varying cell sizes. A more precise phrasing could be:
"Similarly, for Case 2, finer cells with dimensions of 0.05 m × 0.05 m × 0.04 m were used near the fire source, while coarser cells of up to 0.4 m × 0.4 m × 0.3 m were applied in the far-field regions."

Author Response

Please see the attachment

Author Response File: Author Response.docx