Light-Curve Classification of Resident Space Objects for Space Situational Awareness: A Scoping Review

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper reports a scoping review on light-curve classification of Resident Space Objects (RSO) for Space Situational Awareness (SSA). The paper presents a review of AI based method to classify light-curve data. The results of the analysis conducted on three publicly accessible databases are quite interesting and they show a lack of class balance between the different databases in terms of category of the detected RSOs as well as a mismatch of performance between the simulated data used for training and the experimental observations.

I found this paper very informative. The bibliography is rich and appropriate. The paper is well written and contains enough new information and is essentially ready for publication.

I have just a few small minor comments:

• Line 34: I think ‘.’ is missing between ‘environment SSA’ .
• Line 203: It would be beneficial have in an appendix the list of the 25 studies which were included in the final analysis.
• Line 290: I would explicitly write in a note what ‘CIS country’ stay for.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

A significant part of the presented article describes how the authors selected publications for further evaluation in accordance with the PRISMA-ScR guideline. This material closely resembles the introduction to a dissertation. Since the authors did not focus heavily on processing light curves to estimate any of the space object parameters they mentioned, I consider the inclusion of seven pages of text on the selection principles to be inappropriate. Overall, the rest of the presented material may be of interest to specialists dealing with the problems of space debris, its classification, and characterization. I recommend that the authors pay attention to the following minor revisions needed for the text:

1. «Optical light curves, which capture temporal brightness variations, encode information about object type, attitude state, viewing geometry, and surface properties. This paper presents a systematic scoping review of machine learning (ML) and deep learning (DL) methods for RSO classification using light-curve data». Optical light curve DOES NOT encode information about object type. Optical light curve CAN GIVE information about attitude VARIATION and surface properties after procession based on the known data.
2. «Geosynchronous Orbit (GEO)». Typically, GEO means Geostationary Orbit, and Geosynchronous Orbit is designated as GSO.
3. Figure 1, Tables 1-4. Please, decipher the abbreviations (RNN, CNN, PSF, LSTM, and others). It would be helpful to add the Abbreviation Section to the article.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript provides a generally well-structured scoping review of ML/DL approaches for Resident Space Objects (RSO) light-curve classification. The topic is relevant for the SSA community, and the synthesis of datasets, task taxonomies, and methodological trends is clear.

My comments are relatively limited in scope and mainly concern methodological transparency, bibliographic completeness, and a few citation and formatting issues.

Comments

• It would be helpful to explicitly cite the core PRISMA-ScR paper (Tricco et al., 2018), and spell out more clearly how the manuscript actually follows the PRISMA-ScR checklist. For instance, the review should report the exact search strings used for each database, the specific dates when the searches were run, and explain more transparently how non-journal material (e.g., theses or other grey literature) was handled and justified within the inclusion criteria.
• A few peer-reviewed papers within the paper’s scope appear to be missing:

1. 1. Li et al. 2025 (doi: https://doi.org/10.1016/j.actaastro.2025.12.001)
   2. Trummer et al. 2025 (doi: https://doi.org/10.1016/j.actaastro.2024.10.066)
   3. Lu & Zhao 2021, doi: https://doi.org/10.1016/j.chinastron.2021.05.005).

• When general ML methods are introduced, the foundational paper that originally proposed the method should be cited. Domain-specific SSA papers can then be cited as examples of application. For instance, if MAML is discussed as a few-shot/meta-learning framework, the original paper by Finn et al. (ICML 2017) should be cited alongside any SSA application. Similarly, if Barlow Twins is presented as a self-supervised learning method, the canonical paper by Zbontar et al. (ICML 2021) should be referenced, with SSA studies cited separately as implementations in this domain. In its current form the attribution sometimes links methods directly to SSA application papers without referencing the foundational algorithmic source.

• Correct the doi url in Ref. 27.
• 5: incomplete reference.
• missing punctuation in the Introduction (“…sustainability of the orbital environment SSA relies…” needs a period after “environment”)
• missing space in “configurations.The”, line 287.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

This manuscript presents a timely and well-structured systematic scoping review of machine learning and deep learning methods applied to Resident Space Object (RSO) classification using optical light curves. The authors have conducted a thorough literature search, provided a clear methodological approach, and critically assessed publicly available datasets. The paper is well-written and makes a valuable contribution to the field of Space Situational Awareness (SSA). However, several revisions are recommended.

1. The Introduction section could provide a more systematic review of recent advances. Please supplement with the latest progress. The following references are suggested to be included to complete the literature review.
[1] Multi-objective early warning mission planning by multiple satellites using a critical task aggregation-based NSGA-II algorithm[J]. Advances in Space Research, 2025: S027311772501244X.

[2]Analytical Sensitivity Matrix for Near-Optimal Solution to Elliptical Orbit Transfer, IEEE Transactions on Aerospace and Electronic Systems, vol. 62, pp. 461-478, 2026.

2. The review summarizes performance metrics from 25 studies but stops short of any quantitative synthesis. Given the heterogeneity in datasets and evaluation protocols, the authors correctly avoid direct meta-analysis. However, a more structured comparison—such as a table grouping studies—would improve readability and help readers identify trends. The current Tables 3 and 4 are dense and could be better organized with clearer separation between simulated and real-data results.

3. The manuscript correctly identifies class imbalance and repeated observations as challenges, but it does not deeply explore how these biases affect reported accuracy. For instance, many studies report accuracy without accounting for object-level vs. track-level splits, leading to overly optimistic generalization estimates.

4. In Section 4.4, the authors introduce the terms "actionable datasets" and "benchmark datasets" but do not fully define them or explain why existing archives fail to meet either standard. A clearer definition and a checklist of requirements for a benchmark dataset would help guide future work. 

5. Some figure captions (e.g., Figure 2) are incomplete or poorly formatted.

6. Acronyms should be spelled out at first use in each section (e.g., CNN, LSTM, SSA are used repeatedly without reintroduction).

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I thank the authors to have taken into account my comments. From my point of view the paper is ready to be published.

Reviewer 2 Report

Comments and Suggestions for Authors

Thank you for the changes.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors responded satisfactorily to my comments.

Reviewer 4 Report

Comments and Suggestions for Authors

My questions have all been answered.