Deepalaxmi Rajagopal
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Deepalaxmi Rajagopal has a total of 26 co-authors. Here are the top co-authors by number of shared publications:
- Arnd Koeppe
Karlsruhe Institute of Technology, Straße am Forum 7, Karlsruhe, Germany, 76313
4 shared publications - Britta Nestler
Institute for Applied Materials—Microstructure Modelling and Simulation (IAM-MMS), Karlsruhe Institute of Technology (KIT) 1 , Strasse am Forum 7, 76131 Karlsruhe
3 shared publications - Michael Selzer
Karlsruhe Institute of Technology (KIT)
3 shared publications - Adrian Cierpka
Institute of Applied Materials (IAM‐MMS) Karlsruhe Institute of Technology (KIT) Straße am Forum 7 76131 Karlsruhe Germany
2 shared publications - Andreas Hofmann
Institute for Applied Materials (IAM) Karlsruhe Institute of Technology (KIT) Hermann-von-Helmholtz-Platz 1 76344 Eggenstein-Leopoldshafen Germany
2 shared publications - Anna Smith
Institute for Applied Materials (IAM), Karlsruhe Institute of Technology (KIT), Hermann-von-Helmholtz-Platz 1, 76344 Eggenstein-Leopoldshafen, Germany
2 shared publications
Latest Publications
- Open Access
- 8 September 2025
- Published by Wiley in Batteries & Supercaps
- Vol. 9 (2)
Generative Deep Learning for Advanced Battery Materials
The development of battery materials presents a complex multiscale challenge, where optimizing the properties of battery systems across various length scales is essential for achieving targeted performance, enhanced safety, lower costs, and resource availability. Traditional methods for solving this complex, multiscale problem rely on time‐intensive trial‐and‐error approaches, which hinder progress. However, integrating advanced machine learning (ML) frameworks significantly changes the landscape of battery materials research by enabling faster discovery, predictive modeling, and optimization of material properties. Among the ML frameworks, generative deep learning (DL) models stand out, as they capture the statistics of real‐world scenarios by learning an underlying condensed representation of a higher‐dimensional input space to generate information‐rich outputs. By merging computational techniques with experimental research, generative DL provides a significant paradigm shift in analyzing battery materials. This review aims to provide valuable insights into generative models, highlighting their potential to accelerate the characterization, screening, and design of battery materials.
- 8 September 2025
- Published by Wiley in Batteries & Supercaps
- Vol. 9 (2)
Open AccessGenerative Deep Learning for Advanced Battery Materials
The development of battery materials presents a complex multiscale challenge, where optimizing the properties of battery systems across various length scales is essential for achieving targeted performance, enhanced safety, lower costs, and resource availability. Traditional methods for solving this complex, multiscale problem rely on time‐intensive trial‐and‐error approaches, which hinder progress. However, integrating advanced machine learning (ML) frameworks significantly changes the landscape of battery materials research by enabling faster discovery, predictive modeling, and optimization of material properties. Among the ML frameworks, generative deep learning (DL) models stand out, as they capture the statistics of real‐world scenarios by learning an underlying condensed representation of a higher‐dimensional input space to generate information‐rich outputs. By merging computational techniques with experimental research, generative DL provides a significant paradigm shift in analyzing battery materials. This review aims to provide valuable insights into generative models, highlighting their potential to accelerate the characterization, screening, and design of battery materials.
- Open Access
- 16 August 2025
- Published by Wiley in Batteries & Supercaps
- Vol. 8 (12)
Latent Diffusion Models for Virtual Battery Material Screening and Characterization
Efficient characterization of battery materials is fundamental to understanding the underlying electrochemical mechanisms and ensuring the safe operation of batteries. In this work, an innovative data‐driven multimodal generative method is proposed to accelerate the characterization and screening of battery materials. This approach leverages a variant of the latent diffusion model, which combines a variational autoencoder (VAE) and a denoising U‐Net. The VAE maps microscale information from characterization techniques, such as atomic force microscopy (AFM), into a common latent space, and the denoising U‐net, conditioned on battery properties, guides the screening of battery materials. Together, the data‐driven properties of material space, enriched with battery functional properties and formulated in a common latent space, achieve the accurate translation of information from AFM to meaningful material descriptors and accelerate the screening of battery materials to meet the functional needs of the battery system under consideration.
- 16 August 2025
- Published by Wiley in Batteries & Supercaps
- Vol. 8 (12)
Open AccessLatent Diffusion Models for Virtual Battery Material Screening and Characterization
Efficient characterization of battery materials is fundamental to understanding the underlying electrochemical mechanisms and ensuring the safe operation of batteries. In this work, an innovative data‐driven multimodal generative method is proposed to accelerate the characterization and screening of battery materials. This approach leverages a variant of the latent diffusion model, which combines a variational autoencoder (VAE) and a denoising U‐Net. The VAE maps microscale information from characterization techniques, such as atomic force microscopy (AFM), into a common latent space, and the denoising U‐net, conditioned on battery properties, guides the screening of battery materials. Together, the data‐driven properties of material space, enriched with battery functional properties and formulated in a common latent space, achieve the accurate translation of information from AFM to meaningful material descriptors and accelerate the screening of battery materials to meet the functional needs of the battery system under consideration.
- Top Cited PapersOpen Access
- 30 January 2025
- Published by PeerJ in PeerJ Computer Science
- Vol. 11, e2431
AI augmented edge and fog computing for Internet of Health Things (IoHT)
Patients today seek a more advanced and personalized health-care system that keeps up with the pace of modern living. Cloud computing delivers resources over the Internet and enables the deployment of an infinite number of applications to provide services to many sectors. The primary limitation of these cloud frameworks right now is their limited scalability, which results in their inability to meet needs. An edge/fog computing environment, paired with current computing techniques, is the answer to fulfill the energy efficiency and latency requirements for the real-time collection and analysis of health data. Additionally, the Internet of Things (IoT) revolution has been essential in changing contemporary healthcare systems by integrating social, economic, and technological perspectives. This requires transitioning from unadventurous healthcare systems to more adapted healthcare systems that allow patients to be identified, managed, and evaluated more easily. These techniques allow data from many sources to be integrated to effectively assess patient health status and predict potential preventive actions. A subset of the Internet of Things, the Internet of Health Things (IoHT) enables the remote exchange of data for physical processes like patient monitoring, treatment progress, observation, and consultation. Previous surveys related to healthcare mainly focused on architecture and networking, which left untouched important aspects of smart systems like optimal computing techniques such as artificial intelligence, deep learning, advanced technologies, and services that includes 5G and unified communication as a service (UCaaS). This study aims to examine future and existing fog and edge computing architectures and methods that have been augmented with artificial intelligence (AI) for use in healthcare applications, as well as defining the demands and challenges of incorporating fog and edge computing technology in IoHT, thereby helping healthcare professionals and technicians identify the relevant technologies required based on their need for developing IoHT frameworks for remote healthcare. Among the crucial elements to take into account in an IoHT framework are efficient resource management, low latency, and strong security. This review addresses several machine learning techniques for efficient resource management in the IoT, where machine learning (ML) and AI are crucial. It has been noted how the use of modern technologies, such as narrow band-IoT (NB-IoT) for wider coverage and Blockchain technology for security, is transforming IoHT. The last part of the review focuses on the future challenges posed by advanced technologies and services. This study provides prospective research suggestions for enhancing edge and fog computing services for healthcare with modern technologies in order to give patients with an improved quality of life.
- 30 January 2025
- Published by PeerJ in PeerJ Computer Science
- Vol. 11, e2431
Top Cited PapersOpen AccessAI augmented edge and fog computing for Internet of Health Things (IoHT)
Patients today seek a more advanced and personalized health-care system that keeps up with the pace of modern living. Cloud computing delivers resources over the Internet and enables the deployment of an infinite number of applications to provide services to many sectors. The primary limitation of these cloud frameworks right now is their limited scalability, which results in their inability to meet needs. An edge/fog computing environment, paired with current computing techniques, is the answer to fulfill the energy efficiency and latency requirements for the real-time collection and analysis of health data. Additionally, the Internet of Things (IoT) revolution has been essential in changing contemporary healthcare systems by integrating social, economic, and technological perspectives. This requires transitioning from unadventurous healthcare systems to more adapted healthcare systems that allow patients to be identified, managed, and evaluated more easily. These techniques allow data from many sources to be integrated to effectively assess patient health status and predict potential preventive actions. A subset of the Internet of Things, the Internet of Health Things (IoHT) enables the remote exchange of data for physical processes like patient monitoring, treatment progress, observation, and consultation. Previous surveys related to healthcare mainly focused on architecture and networking, which left untouched important aspects of smart systems like optimal computing techniques such as artificial intelligence, deep learning, advanced technologies, and services that includes 5G and unified communication as a service (UCaaS). This study aims to examine future and existing fog and edge computing architectures and methods that have been augmented with artificial intelligence (AI) for use in healthcare applications, as well as defining the demands and challenges of incorporating fog and edge computing technology in IoHT, thereby helping healthcare professionals and technicians identify the relevant technologies required based on their need for developing IoHT frameworks for remote healthcare. Among the crucial elements to take into account in an IoHT framework are efficient resource management, low latency, and strong security. This review addresses several machine learning techniques for efficient resource management in the IoT, where machine learning (ML) and AI are crucial. It has been noted how the use of modern technologies, such as narrow band-IoT (NB-IoT) for wider coverage and Blockchain technology for security, is transforming IoHT. The last part of the review focuses on the future challenges posed by advanced technologies and services. This study provides prospective research suggestions for enhancing edge and fog computing services for healthcare with modern technologies in order to give patients with an improved quality of life.
- Open Access
- 6 September 2024
- Published by Wiley in Batteries & Supercaps
- Vol. 7 (12) , e202400406
From Powder to Pouch Cell: Setting up a Sodium‐Ion Battery Reference System Based on Na 3 V 2 (PO 4 ) 3 /C and Hard Carbon
At the research level, novel active materials for batteries are synthesised on a small scale, fabricated into electrodes and electrochemically characterised using each group's established process due to the lack of standards. Recently, eminent researchers have criticised the implementation of e. g. low active material contents/electrode loadings, the use of research‐type battery cell constructions, or the lack of statistically relevant data, resulting in overstated data and thus giving misleading predictions of the key performance indicators of new battery technologies. Here, we report on the establishment of a reference system for the development of sodium‐ion batteries. Electrodes are fabricated under relevant conditions using 9.5 mg/cm 2 self‐synthesised Na 3 V 2 (PO 4 ) 3 /C cathode active material and 3.6 mg/cm 2 commercially available hard carbon anode active material. It is found that different types of battery cells are more or less suitable for half‐ and/or full‐cell testing, resulting in ir/reproducible or underestimated active material capacities. Furthermore, the influence of electrode overhang, which is relevant for upscaling, is evaluated. The demonstrator cell (TRL 4–5) has been further characterised providing measured data on the power/energy density and thermal behaviour during rate testing up to 15 C and projections are made for its practical limits.
- 6 September 2024
- Published by Wiley in Batteries & Supercaps
- Vol. 7 (12) , e202400406
Open AccessFrom Powder to Pouch Cell: Setting up a Sodium‐Ion Battery Reference System Based on Na 3 V 2 (PO 4 ) 3 /C and Hard Carbon
At the research level, novel active materials for batteries are synthesised on a small scale, fabricated into electrodes and electrochemically characterised using each group's established process due to the lack of standards. Recently, eminent researchers have criticised the implementation of e. g. low active material contents/electrode loadings, the use of research‐type battery cell constructions, or the lack of statistically relevant data, resulting in overstated data and thus giving misleading predictions of the key performance indicators of new battery technologies. Here, we report on the establishment of a reference system for the development of sodium‐ion batteries. Electrodes are fabricated under relevant conditions using 9.5 mg/cm 2 self‐synthesised Na 3 V 2 (PO 4 ) 3 /C cathode active material and 3.6 mg/cm 2 commercially available hard carbon anode active material. It is found that different types of battery cells are more or less suitable for half‐ and/or full‐cell testing, resulting in ir/reproducible or underestimated active material capacities. Furthermore, the influence of electrode overhang, which is relevant for upscaling, is evaluated. The demonstrator cell (TRL 4–5) has been further characterised providing measured data on the power/energy density and thermal behaviour during rate testing up to 15 C and projections are made for its practical limits.
- Open Access
- 21 June 2024
- Published by American Chemical Society (ACS) in ChemRxiv
From Powder to Pouch Cell: Setting up a Sodium-Ion Battery Reference System Based on Na3V2(PO4)3/C and Hard Carbon
At the research level, novel active materials for batteries are synthesised on a small scale, fabricated into electrodes and electrochemically characterised using each group’s established process due to the lack of standards. Recently, eminent researchers have criticised the implementation of e.g. low active material contents/electrode loadings, the use of research-type battery cell constructions, or the lack of statistically relevant data, resulting in overstated data and thus giving misleading predictions of the key performance indicators of new battery technologies. Here, we report on the establishment of a reference system for the development of sodium-ion batteries. Electrodes are fabricated under relevant conditions using 9.5 mg/cm² self-synthesised Na3V2(PO4)3/C cathode active material and 3.6 mg/cm² commercially available hard carbon anode active material. It is found that different types of battery cells are more or less suitable for half- and/ or full-cell testing, resulting in ir/reproducible or underestimated active material capacities. Furthermore, the influence of electrode overhang, which is relevant for upscaling, is evaluated. The demonstrator cell (TRL 4-5) has been further characterised providing measured data on the power/energy density and thermal behaviour during rate testing up to 15 C and projections are made for its practical limits.
- 21 June 2024
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessFrom Powder to Pouch Cell: Setting up a Sodium-Ion Battery Reference System Based on Na3V2(PO4)3/C and Hard Carbon
At the research level, novel active materials for batteries are synthesised on a small scale, fabricated into electrodes and electrochemically characterised using each group’s established process due to the lack of standards. Recently, eminent researchers have criticised the implementation of e.g. low active material contents/electrode loadings, the use of research-type battery cell constructions, or the lack of statistically relevant data, resulting in overstated data and thus giving misleading predictions of the key performance indicators of new battery technologies. Here, we report on the establishment of a reference system for the development of sodium-ion batteries. Electrodes are fabricated under relevant conditions using 9.5 mg/cm² self-synthesised Na3V2(PO4)3/C cathode active material and 3.6 mg/cm² commercially available hard carbon anode active material. It is found that different types of battery cells are more or less suitable for half- and/ or full-cell testing, resulting in ir/reproducible or underestimated active material capacities. Furthermore, the influence of electrode overhang, which is relevant for upscaling, is evaluated. The demonstrator cell (TRL 4-5) has been further characterised providing measured data on the power/energy density and thermal behaviour during rate testing up to 15 C and projections are made for its practical limits.
- Open Access
- 12 September 2023
- Published by Wiley in Advanced Energy Materials
- Vol. 13 (40)
Data‐Driven Virtual Material Analysis and Synthesis for Solid Electrolyte Interphases
Solid electrolyte interphases (SEIs) form as reduction products at the electrodes and strongly affect battery performance and safety. Because SEI formation poses a highly nonlinear, complex multi‐physics problem over various lengths and time scales, traditional modeling approaches struggle to characterize SEI evolution solely with existing physical properties. To improve the characterization of SEIs, it proposes a data‐driven strategy for a virtual material design that learns to represent and characterize SEI formation with physical and data‐driven properties from kinetic Monte Carlo simulations. A Variational AutoEncoder with a property regressor learns data‐driven properties, which represent SEI configurations and correlate with physical target properties. This new neural network design encodes the high‐dimensional structural and reaction spaces into a lower‐dimensional latent space, while the property regressor orders the latent space by physical target properties. The model achieves high correlation scores between target and predicted properties from latent representations, thereby proving that the data‐driven properties enrich the expressiveness of SEI characterizations.
- 12 September 2023
- Published by Wiley in Advanced Energy Materials
- Vol. 13 (40)
Open AccessData‐Driven Virtual Material Analysis and Synthesis for Solid Electrolyte Interphases
Solid electrolyte interphases (SEIs) form as reduction products at the electrodes and strongly affect battery performance and safety. Because SEI formation poses a highly nonlinear, complex multi‐physics problem over various lengths and time scales, traditional modeling approaches struggle to characterize SEI evolution solely with existing physical properties. To improve the characterization of SEIs, it proposes a data‐driven strategy for a virtual material design that learns to represent and characterize SEI formation with physical and data‐driven properties from kinetic Monte Carlo simulations. A Variational AutoEncoder with a property regressor learns data‐driven properties, which represent SEI configurations and correlate with physical target properties. This new neural network design encodes the high‐dimensional structural and reaction spaces into a lower‐dimensional latent space, while the property regressor orders the latent space by physical target properties. The model achieves high correlation scores between target and predicted properties from latent representations, thereby proving that the data‐driven properties enrich the expressiveness of SEI characterizations.