Applications and Challenges of Artificial Intelligence in Space Missions
Open accessThe DOI matches its Crossref record, with a small difference in title or year; 256 of the 268 references it cites were found in a registry or an index; 29 of the 213 sources its claims cite were checked for support.
Abstract
Artificial Intelligence (AI) is increasingly finding acceptance in the space community, especially Machine Learning (ML), a subdomain of AI. ML algorithms now find numerous applications in autonomous navigation, spacecraft health monitoring and operational management of satellite constellations. However, a large number of surveys on the applications of AI in space missions can be classified into two categories. The first category suffers from the limitation of being old and not covering some crucial and recent developments in the field; such as the contributions of Deep Learning (DL) and bioinspired AI algorithms. The challenge with the second category lies in its being too detailed with respect to the development and application of specific AI techniques or algorithms. These limitations have necessitated the need to have a concise survey with a wider scope for those interested in the applications and challenges of AI in the space industry, especially those with technical backgrounds in other fields. In this paper, we surveyed the use of traditional AI techniques in various domains of space missions without delving into formal methods. Some bioinspired AI algorithms were also surveyed and their potential application areas highlighted. Unlike similar surveys that focus only on technological challenges, we also addressed some crucial legal drawbacks that emanate from the reliance and use of AI in space. Summarily, while discussing future directions we reviewed some advancements in Internet of Things (IoT) and Blockchain technologies. Our review prioritized three application domains positioned to benefit immensely from the inevitable AI-Blockchain convergence in the space community. These include the Internet of Space Things (IoST), Satellite Communication and Spacecraft Data Security.
Overview
CoChat Summary
TL;DR
This survey reviews AI and machine learning applications across space missions, including spacecraft health monitoring, remote sensing, satellite communications, and robotic autonomous systems, with attention to bioinspired algorithms and legal challenges. It argues AI can increase autonomy and data return but faces radiation, computational, telemetry, and regulatory hurdles.
Key findings
- AI/ML are applied in spacecraft health monitoring (e.g., SHARP; data-mining tools Orca1 and IMS; LSTMs on Mars Science Laboratory Curiosity and SMAP), remote sensing (DL for on-board preprocessing and cloud detection; SAR), satellite communications (VHTS targeting ~1 Terabit/s; CNNs for flexible payload DRM), and robotic autonomous systems (SLAM, perception, learning).
- A PCA + multi-class SVM fault diagnosis approach achieved 99.2% and 97.4% classification accuracy in Case 1 and Case 2, respectively; ANN achieved 99.13% and 92.6%. SVM training times were 36.2 s and 19.5 s, and test times 11.2 s and 5.8 s.
- A radiation study found k-means could withstand radiation in Earth orbit without radiation-hardened memory; simpler algorithms (regular k-means, linear SVM) were less sensitive than more complex versions (kd-k-means, Gaussian SVM), and kd-k-means was not recommended for onboard spacecraft use.
- The paper reports that a 9% satellite operational failure rate plus a 4–5% launch vehicle failure rate implies approximately one in seven satellites fails before end of life. EO-1’s Autonomous Sciencecraft Experiment cut operational costs by $1M per year with a 50% increase in science return. Legally, no international or space treaties address AI in space, so domestic legislation must serve as the principal source of substantive law.
Methods
- Narrative survey of AI in space missions covering four application areas: Spacecraft Health Monitoring, Remote Sensing, Satellite Communications, and Robotic Autonomous Systems for space; it also reviews ML categories, autonomy, bioinspired algorithms, and legal challenges.
- Draws on published literature and mission examples (e.g., NASA/JPL SHARP, Orca1, IMS, MSL Curiosity, SMAP, EO-1, ISS, Mars rovers); no primary data collection or experiments are reported.
- Analysis is qualitative synthesis; the authors state they exclude complex mathematical formulas and abstractions and focus more on communication and remote sensing satellites.
Limitations
- The paper states that prior surveys are either old and miss recent DL and bioinspired AI developments, or are too detailed on specific techniques; this survey aims for a concise wider scope but excludes formal methods and complex mathematics.
- The paper identifies AI-in-space limitations including radiation effects on electronics, computational constraints and legacy processors, heterogeneous multimodal telemetry, lack of labeled anomalies, and legal gaps such as the absence of international AI-in-space treaties.
- Evident from the text: as a narrative survey, it reports no primary experiments or systematic review protocol, and several sections are brief or rely on cited secondary sources.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
17 Valid · 31 Not valid · 158 Not checked · 101 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Summarily, most studies surveying AI with respect to space applications have either been reviews concerning closely related aspects of space missions (([5], [7]) or detailed topical reviews with respect to the development and application of specific AI techniques or algorithms ([15], [11] and [16]) to a specific field ([17]).”
What the sources say
- Not matchedcites [5] Meß, Techniques of artificial intelligence for space applications—A survey
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Partly supportedcites [7] sourceExperimental — accuracy not yet measured
The Introduction (I. INTRODUCTION) of Kunze et al. (2018) makes a structurally parallel point but scoped to long-term autonomy in general, not to space applications: existing special issues and surveys 'relate to LTA' by covering AI methods within integrated robot systems [3], [4], or are topical collections on specific techniques within a specific field (perception: localisation and mapping [5], [6], calibration [7], visual place recognition [8]), and 'this is the first survey that focuses specifically on AI techniques for enabling long-term robot autonomy.' Nothing in our copy characterises the literature of surveys of AI for space applications as such; the space section (Domain Application Features Duration AI Areas System) discusses deployed systems (Opportunity/MAPGEN [9], stereo navigation [10], IPEX [11]) rather than reviewing survey papers about space. Thus the 'closely related aspects / topical reviews of specific AI techniques to a specific field' dichotomy is supported, but the transfer of it to 'most studies surveying AI with respect to space applications' is not stated by this source. (LLM, conf=0.65)
- Could not verifycites [15] sourceExperimental — accuracy not yet measured
Our copy of Kothari, Liberis & Lane (2020) describes only its own scope — the Abstract (p. 1) says they 'collate various applications of machine learning to space data, such as satellite imaging', and the introduction/summary (p. 1) that they 'summarise prominent applications of ML in space'; Sec. 2 (pp. 1-2) then surveys payload-data and spacecraft-operation work. Nowhere does it characterise the existing survey literature on AI in space as consisting mostly of reviews of closely related aspects of space missions or of topical reviews of specific AI techniques/algorithms applied to one field. I searched for prior surveys/reviews, related work, limitations of existing overviews, and 'survey'/'review' phrasings; the only review-type items in our copy are individual references (e.g. [35] Paoletti et al. 2019 on hyperspectral DL classifiers, [40] Sze et al. 2017) with no such framing statement. Nothing contrary is said either, so the claim is not settled by this source. (LLM, conf=0.65)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Not matched
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Traditional AI algorithms are usually met with developmental constraints such as the reliance on high-precision sensors and complex computing [54] in this field.”
What the sources say
- Partly supportedcites [54] sourceExperimental — accuracy not yet measured
The Introduction says traditional AI methods have limitations and that bioinspired methods are usually more efficient. Future Directions discusses the difficulty and cost of improving traditional sensor precision. These passages do not establish that traditional AI algorithms generally rely on high-precision sensors or complex computing in mobile robot control. (LLM, conf=0.82)
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“For example, by using the Autonomous Sciencecraft Experiment to automatically plan and adaptively re-plan observations as required the EO-1 spacecraft was able to cut operational costs by $1 M per year, with a 50% increase in science return [118].”
What the sources say
- Supportedcites [118] source
Section 3, “Opportunities for machine learning in space” (page not provided), states that EO-1 reduced operational costs by $1 million per year and increased science return by 50% using the Autonomous Sciencecraft Experiment to plan and adaptively re-plan observations. (LLM, conf=0.95)
plainstates a causeSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“Parts of the image of no interest (e.g. those occluded by clouds) can be discarded so as to drastically reduce the amount of data transmitted [139],[15].”
What the sources say
- Not checked — no full text of the sourcecites [139] source
- Supportedcites [15] source
Section 4.1 'Efficient Satellite Imaging' (p. 3) of our copy states that on-board pre-processing with deep learning 'can discard parts of the image of no interest, e.g. occluded by clouds', and, citing global annual cloud coverage of ~66%, that 'excluding cloud images would drastically reduce the amount of data transmitted'. The citing sentence reproduces both the mechanism and the qualifier (cloud occlusion) faithfully. (LLM, conf=0.92)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“This implies that a lack of labeled anomalies necessitates the use of unsupervised or semi-supervised approaches [134].”
What the sources say
- Partly supportedcites [134] sourceExperimental — accuracy not yet measured
Our copy supports the rationale only implicitly: IV. Summary and Future Work treats supervised learning (decision tree / SVM) as a future application contingent on whether 'examples of fault behavior are available in the archived data,' i.e. labeled fault data is generally lacking, while the Abstract and Sections A–C show the paper's actual methods are unsupervised (Orca nearest-neighbor outlier detection, IMS clustering of nominal operating regions on unlabeled telemetry archives). However, the paper never states that this scarcity 'necessitates' unsupervised methods, never uses the terms 'labeled' or 'semi-supervised,' and says nothing about semi-supervised approaches at all, so that part of the attributed claim is not in our copy. (LLM, conf=0.68)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
107 more claims · 15 with their cited sources checked
Create a free account to see all 307 claims and to verify the 15 references still open.
- It may state two separate claims as one. 64
- Probably not a substantial claim. 164
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ A. ANOMALY DETECTION & FAULT DETECTION
“FIGURE 5. NigComSat 1 (Courtesy of NASRDA)”
§ A. ANOMALY DETECTION & FAULT DETECTION
“Fig. 5 shows Nigerian Sat1, which was lost because of a failure that prevented one of the solar arrays from deploying after launch.”
Citations
Cites
Reinforcement Learning: An Introduction
Sutton, R. S., & Barto, A. G. (1998) · IEEE Transactions on Neural Networks · doi:10.1109/tnn.1998.712192 · Cited by 27,225 · cited 1×
“An RL agent has the goal of learning the best way to accomplish a task through repeated interactions with its environment [44].”
BackgroundMirjalili, S., Mirjalili, S. M., & Lewis, A. (2014) · Advances in Engineering Software · doi:10.1016/j.advengsoft.2013.12.007 · Cited by 19,343
No free lunch theorems for optimization
Wolpert, D. H., & Macready, W. G. (1997) · IEEE Transactions on Evolutionary Computation · doi:10.1109/4235.585893 · Cited by 14,237 · cited 1×
“According to the No Free Lunch (NFL) theorem [69], the superior performance of an optimizer on a class of applications or problems cannot guarantee a similar performance on other problems.”
BackgroundArtificial intelligence: A modern approach
Nilsson, N. J. (1996) · Artificial Intelligence · doi:10.1016/0004-3702(96)00007-0 · Cited by 10,583
Principal component analysis: A review and recent developments
Jolliffe, I. T., & Cadima, J. (2016) · Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences · doi:10.1098/rsta.2015.0202 · Cited by 10,117 · cited 1×
“statistical information) as possible [43].”
MOEA/D: A Multiobjective Evolutionary Algorithm Based on Decomposition
Zhang, Q., & Li, H. (2007) · IEEE Transactions on Evolutionary Computation · doi:10.1109/tevc.2007.892759 · Cited by 10,069 · cited 1×
“It has been shown that evolutionary algorithms such as Non-Dominated Sorting Genetic Algorithm (NSGA-I) [73] and Multi-Objective Evolutionary Algorithm based on Decomposition (MOEA/D) [74], can approximate the true PO solutions of multi-objective problems [70].”
MethodROS: an open-source Robot Operating System
Quigley, M. (2009) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · Cited by 8,729
Neural Networks And Learning Machines
Haykin, S. (2010) · Cited by 7,534
Muiltiobjective Optimization Using Nondominated Sorting in Genetic Algorithms
Srinivas, N., & Deb, K. (1994) · Evolutionary Computation · doi:10.1162/evco.1994.2.3.221 · Cited by 6,935 · cited 1×
“It has been shown that evolutionary algorithms such as Non-Dominated Sorting Genetic Algorithm (NSGA-I) [73] and Multi-Objective Evolutionary Algorithm based on Decomposition (MOEA/D) [74], can approximate the true PO solutions of multi-objective problems [70].”
MethodMoth-flame optimization algorithm: A novel nature-inspired heuristic paradigm
Mirjalili, S. (2015) · Knowledge-Based Systems · doi:10.1016/j.knosys.2015.07.006 · Cited by 4,626
Research works printed but not linked: 26
- 2010, Challenger Explosion
- Laura, 2018, Russia Investigating Failed Soyuz Rocket Launch
- Prosser, 2018, AI is kicking space exploration into hyper drive—Here’s how?
- Mitchell, 1997, Machine Learning
- Burkov, 2019, The Hundred-Page Machine Learning Book
- Yin, 2012, Improved invasive weed optimization based on hybrid genetic algorithm, J. Comput. Inf. Syst., 8, 3437
- Yuso, 2014, Artificial immune system for flexible manufacturing system machine loading problem, ICIC Exp. Lett., 8, 709
- Reynolds, An introduction to cultural algorithms, 131
- Wu, 2014, An efficient image matching algorithm based on culture evolution, J. Chem. Pharmaceutical Res., 6, 271
- Cui, 2016, Projection pursuit model for evaluation of flood and drought disasters based on chicken swarm optimization algorithm, Adv. Sci. Technol. Water Resour., 36, 5
- Skelly, 2021, Use case: Autonomy for space systems
- Frost, Challenges and opportunities for autonomous systems in space, 1
- Lawal, Optimization of mass volume ratio of spacecraft structure for advanced and high powered communication satellites
- Robinson, Applying model-based reasoning to the FDIR of the command and data handling subsystem of the ISS
- Komissarov, 2020, AI applications for satellite imagery and satellite data
- Soenen, 2019, Deep Learning and SAR Applications: A Short Overview of Advancements and Challenges in Earth Observation Applications, Medium
- Montemerlo, FastSLAM 2.0: An improved particle filtering algorithm for simultaneous localization and mapping that provably converges, 1151
- Clark, 2021, Space Station’s Dextre Robot Passes Crucial Test
- 2020, 2017 Space Launch Statistics
- Wadhwani, 2021, Space Robotics Market Size by Solution (Product [Rovers/Landers, Probes, Arms/Manipulator Systems], Software, Services), by Application (Deep Space [Space Exploration, Space Transportation], Near Space [Space Exploration, Space Transportation]), by End-Use (Commercial, Government), Industry Analysis Report, Regional Outlook, Application Potential, Price Trend, Competitive Market Share & Forecast, 2019–2025
- Cassel, 2021, The Hardware and Software Used in Space
- 2021, Radiation-Hardened Space Electronics Enter the Multi-Core Era
- Gal, 2020, Artificial intelligence in space, arXiv:2006.12362 · arXiv
- Stankovic, 2017, Exploring legal, ethical and policy implications of artificial intelligence
- Carlin, 2020, U.S. department of commerce imposes immediate export controls on artificial intelligence software used to automatically detect and identify objects remotely, J. Robot., Artif. Intell. Law, 3, 1
- O’Neal, 2018, Houston, We Have a Solution: Blockchain in the Space Industry
Cited by
Reinforcement learning in spacecraft control applications: Advances, prospects, and challenges
Tipaldi, M., Iervolino, R., & Massenio, P. R. (2022) · Annual Reviews in Control · doi:10.1016/j.arcontrol.2022.07.004 · Cited by 141
Ong, J., Waisberg, E., Masalkhi, M., Kamran, S. A., Lowry, K., Sarker, P., Zaman, N., Paladugu, P., Tavakkoli, A., & Lee, A. G. (2023) · Brain Sciences · doi:10.3390/brainsci13081148 · Cited by 41
A prospective on machine learning challenges, progress, and potential in polymer science
Struble, D. C., Lamb, B. G., & Ma, B. (2024) · MRS Communications · doi:10.1557/s43579-024-00587-8 · Cited by 48
Metaheuristics role in image processing and computer vision applications: a comprehensive review
Şahin, M. F., & Anka, F. (2025) · Cluster Computing · doi:10.1007/s10586-025-05610-8 · Cited by 34
“In a variety of disciplines, including [23], autonomous systems [24], space [25], and numerous other domains [26–28].”
Computer vision tasks for intelligent aerospace perception: An overview
Chen, H., Sun, Q., Li, F., & Tang, Y. (2024) · Science China Technological Sciences · doi:10.1007/s11431-024-2714-4 · Cited by 30
“…concentrate on particular areas such as robotic manipulation [26], guidance, navigation and control [27–29], debris removal [1], among others [30, 31], a scarcity exists in terms of surveys that o ff er an exhaustive overview of computer vision tasks for aerospace perception, especially…”
Metadata
- Authors
- Oche, P. A., Ewa, G. A., & Ibekwe, N.
- Affiliations
- National Space Research and Development Agency (NG)
- Corresponding author
- Oche, Paul A.
- Venue
- IEEE Access
- Published
- 2024
- Volume
- 12
- Pages
- 44481-44509
- ISSN
- 2169-3536 (ISSN-L)
- OpenAlex
- W4206200704
- Semantic Scholar
- 4cbf87b2cf63458e4f2977799f3315dd844cd26c
- Cited by
- 68
- References
- 268
- Citations by year
- 2026: 13 · 2025: 33 · 2024: 15
- FWCI
- 1.93
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Locations
- IEEE Access (publishedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- National Space Research and Development Agency
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Artificial Intelligence Applications
- Social Sciences › Political Science and International Relations › Space exploration and regulation
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Space Technology and Applications
- Keywords
- machine learning, deep learning, bio-inspired algorithms, autonomous spacecraft navigation, spacecraft health monitoring, blockchain, satellite communication, Internet of Things
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (qwen3.8-omni-flash and gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Applications and Challenges of Artificial Intelligence in Space Missions. https://app.cochat.ai/papers/10.1109/access.2021.3132500
- The paper
- Oche, P. A., Ewa, G. A., & Ibekwe, N. (2024). Applications and Challenges of Artificial Intelligence in Space Missions. IEEE Access. https://doi.org/10.1109/access.2021.3132500
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.