A critical review on the state-of-the-art and future prospects of machine learning for Earth observation operations

Miralles, Pablo; Thangavel, Kathiravan; Scannapieco, Antonio Fulvio; Jagadam, Nitya; Baranwal, Prerna; Faldu, Bhavin; Abhang, Ruchita; Bhatia, Sahil; Bonnart, Sebastien; Bhatnagar, Ishita; Batul, Beenish; Prasad, P. Nagendra; Ortega-González, Héctor; Joseph, Harrish; More, Harshal; Morchedi, Sondes; Panda, Aman Kumar; Fraia, Marco Zaccaria Di; Wischert, Daniel; Stepanova, Daria

doi:10.1016/j.asr.2023.02.025

Cited by 17 publications

(2 citation statements)

References 105 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This provides enhanced performance for time-critical applications such as a rare event, disaster events, etc. With the ISL and the DSS, real-time operations are achieved by adding predictive and reactive elements [ 28 , 29 , 30 ] within the architecture that endows the iDSS.…”

Section: Distributed Satellite Systemsmentioning

confidence: 99%

Trusted Autonomous Operations of Distributed Satellite Systems Using Optical Sensors

Thangavel

Spiller

Sabatini

et al. 2023

Sensors

Self Cite

View full text Add to dashboard Cite

Recent developments in Distributed Satellite Systems (DSS) have undoubtedly increased mission value due to the ability to reconfigure the spacecraft cluster/formation and incrementally add new or update older satellites in the formation. These features provide inherent benefits, such as increased mission effectiveness, multi-mission capabilities, design flexibility, and so on. Trusted Autonomous Satellite Operation (TASO) are possible owing to the predictive and reactive integrity features offered by Artificial Intelligence (AI), including both on-board satellites and in the ground control segments. To effectively monitor and manage time-critical events such as disaster relief missions, the DSS must be able to reconfigure autonomously. To achieve TASO, the DSS should have reconfiguration capability within the architecture and spacecraft should communicate with each other through an Inter-Satellite Link (ISL). Recent advances in AI, sensing, and computing technologies have resulted in the development of new promising concepts for the safe and efficient operation of the DSS. The combination of these technologies enables trusted autonomy in intelligent DSS (iDSS) operations, allowing for a more responsive and resilient approach to Space Mission Management (SMM) in terms of data collection and processing, especially when using state-of-the-art optical sensors. This research looks into the potential applications of iDSS by proposing a constellation of satellites in Low Earth Orbit (LEO) for near-real-time wildfire management. For spacecraft to continuously monitor Areas of Interest (AOI) in a dynamically changing environment, satellite missions must have extensive coverage, revisit intervals, and reconfiguration capability that iDSS can offer. Our recent work demonstrated the feasibility of AI-based data processing using state-of-the-art on-board astrionics hardware accelerators. Based on these initial results, AI-based software has been successively developed for wildfire detection on-board iDSS satellites. To demonstrate the applicability of the proposed iDSS architecture, simulation case studies are performed considering different geographic locations.

show abstract

Section: Distributed Satellite Systemsmentioning

confidence: 99%

Trusted Autonomous Operations of Distributed Satellite Systems Using Optical Sensors

Thangavel

Spiller

Sabatini

et al. 2023

Sensors

Self Cite

View full text Add to dashboard Cite

show abstract

“…The quality of the information that can be extracted from PRISMA HS imagery was investigated in [29], where analytical methodologies were proposed to locate wildfires and estimate the temperature of active fire pixels. At the same time, we showed the possibility of implementing Trusted Autonomous Satellite Operations [30][31][32] by utilizing artificial intelligence [33,34] on-board satellites with astrionics for data processing [35][36][37]. For the purpose of providing real-time or near real-time disaster management, the same has been used in Distributed Satellite Systems [38][39][40].…”

Section: Introductionmentioning

confidence: 99%

Wildfire Detection Using Convolutional Neural Networks and PRISMA Hyperspectral Imagery: A Spatial-Spectral Analysis

Spiller,

Carbone,

Amici

et al. 2023

Remote Sensing

Self Cite

View full text Add to dashboard Cite

The exacerbation of wildfires, attributed to the effects of climate change, presents substantial risks to ecological systems, infrastructure, and human well-being. In the context of the Sustainable Development Goals (SDGs), particularly those related to climate action, prioritizing the assessment and management of the occurrence and intensity of extensive wildfires is of utmost importance. In recent times, there has been a significant increase in the frequency and severity of widespread wildfires worldwide, affecting several locations, including Australia, Italy, and the United States of America. The presence of complex phenomena marked by limited predictability leads to significant negative impacts on biodiversity and human lives. The utilization of satellite-derived data with neural networks, such as convolutional neural networks (CNNs), is a potentially advantageous approach for augmenting the monitoring capabilities of wildfires. This research examines the generalization capability of four neural network models, namely the fully connected (FC), one-dimensional (1D) CNN, two-dimensional (2D) CNN, and three-dimensional (3D) CNN model. Each model’s performance, as measured by accuracy, recall, and F1 scores, is assessed through K-fold cross-validation. Subsequently, T-statistics and p-values are computed based on these metrics to conduct a statistical comparison among the different models, allowing us to quantify the degree of similarity or dissimilarity between them. By using training data from Australia and Sicily, the performances of the trained model are evaluated on the test dataset from Oregon. The results are promising, with cross-validation on the training dataset producing mean precision, recall, and F1 scores ranging between approximately 0.97 and 0.98. Especially, the fully connected model has superior generalization capabilities, whilst the 3D CNN offers more refined and less distorted classifications. However, certain issues, such as false fire detection and confusion between smoke and shadows, persist. The aforementioned methodologies offer significant perspectives on the capabilities of neural network technologies in supporting the detection and management of wildfires. These approaches address the crucial matter of domain transferability and the associated dependability of predictions in new regions. This study makes a valuable contribution to the ongoing efforts in climate change by assisting in monitoring and managing wildfires.

show abstract

Profiling Power Consumption for Deep Learning on Resource Limited Devices

Duggan,

Scully,

Smith

et al. 2023

Lecture Notes in Computer Science

View full text Add to dashboard Cite

A critical review on the state-of-the-art and future prospects of machine learning for Earth observation operations

Cited by 17 publications

References 105 publications

Trusted Autonomous Operations of Distributed Satellite Systems Using Optical Sensors

Trusted Autonomous Operations of Distributed Satellite Systems Using Optical Sensors

Wildfire Detection Using Convolutional Neural Networks and PRISMA Hyperspectral Imagery: A Spatial-Spectral Analysis

Profiling Power Consumption for Deep Learning on Resource Limited Devices

Contact Info

Product

Resources

About