Application-specific Target Research Articles

(Invited) Fuel Cell Technologies: A U.S. Department of Energy Perspective

The U.S. Department of Energy (DOE) supports activities focused on the development of fuel cell technologies that are competitive with incumbent and emerging technologies across diverse applications. Fuel cell technologies for heavy-duty transportation applications are of particular interest as significant reductions in both carbon emissions and criteria pollutant emissions can be achieved.DOE engages in fuel cell research, development, and demonstration (RD&D), working closely with its national laboratories, universities, and industry partners to overcome critical technical barriers to fuel cell development. For heavy-duty applications, RD&D is focused on the development of direct hydrogen-fueled proton-exchange membrane fuel cells (PEMFCs). PEMFC RD&D addresses the materials, component, system, manufacturing, and supply chain challenges to accelerate the commercialization of zero-emission medium- and heavy-duty vehicles, with transferable benefits for other applications.The DOE’s RD&D strategy for fuel cell technologies is driven by application-specific targets. For long-haul fuel cell trucks, these include 2030 targets for cost ($80/kW at high volume production) and durability (25,000 hours). For comparison, the cost in 2023 of a PEMFC system for long-haul trucks based on next-generation laboratory demonstrated technology is projected to be approximately $170/kW when manufactured at a volume of 50,000 units/year.DOE supports a portfolio of RD&D projects and has established the national lab-led Million Mile Fuel Cell Truck Consortium (M2FCT) to advance durability and efficiency of PEMFCs for heavy-duty vehicle applications. M2FCT’s RD&D focuses on meeting a membrane electrode assembly (MEA) target by 2025 that combines efficiency, durability, and cost in a single goal: 2.5 kW/gPGM specific power (1.07 A/cm2 current density) at 0.7 V after 25,000 hour-equivalent accelerated durability test.Supported by the U.S. Infrastructure Investment and Jobs Act, DOE pursues advances in manufacturing and recycling of clean hydrogen technologies to bridge the gap between technology development and deployment. Enabling economies of scale will require a reliable supply of components, automation of the cell and stack assembly, and manufacturing capabilities not seen in the field to date. DOE established capacity targets for heavy-duty fuel cell component and stack production, including a target of 20,000 stacks per year in a single production line by 2030 to enable market lift-off.The DOE’s RD&D portfolio includes recently selected industry-led manufacturing and component supply chain development projects. To complement these projects, DOE relaunched the national lab-led Roll-to-Roll Consortium to advance high-throughput fuel cell and electrolyzer manufacturing, with emphasis on PEM MEA and MEA components.End-of-life challenges for fuel cells and electrolyzers are also addressed. DOE is establishing a consortium of industry, academia, and national labs to develop innovative and practical approaches to enable the recovery, recycling, and reuse of clean hydrogen materials and components, with the goal of securing long-term supply chain security and environmental sustainability.

Analysis of the Decomposition of Sulfur-Based Electrolyte Additives in Spent LiNi0.6Co0.2Mn0.2O2||AG Cells

The development and optimization processes of the last decades in the context of lithium ion batteries (LIBs) have led to a variety of electrode chemistries, which enable application-specific targets such as high energy, high power, high safety or long lifetime. Most of these LIB systems rely on a liquid carbonate-based electrolyte, whose basic composition has changed scarcely during this period.1 In this respect, the usage of electrolyte additives is one of the most economical and effective ways to further improve the overall performance, while also impact the application-tailored characteristics of the battery cells at the elctrolyte level without changing the bulk properties.2 These improvements can range from overcharge protection to lowered flammability or interphase formation, among others. In particular, film-forming additives signifcantly influence the performance and life time of LIBs by preventing excessive decomposition of the electrolyte, which is in contact with the positive and negative electrodes, by the formation of a passivation layer. Investigating the electrochemical decomposition of the electrolyte components, especially electrolyte additives, is key to enable a better understanding of the composition from the electrode||electrolyte interface layer and the effects on LIB cell performance.3 Thus, aging mechanisms and parasitic reactions can be elucidated to improve and design new electrolyte additives.Herein, the irreversible decomposition of sulfur-containing electrolyte additives will be addressed. Sulfur-containing additives represent an attractive option for film-forming additives as these compounds endow lower energies in the lowest unoccupied molecular orbital (LUMO) compared to the organic carbonate analogs and therefore are more susceptible to electrochemical reduction.4 However, these class of electrolyte additives find only a limited use in state-of-the-art electrolyte mixtures due to the often reported or suspected carcinogenicity and toxicity towards the human body. Nevertheless, since sulfur-containing additives still play an important role in spent LIBs of recent decades, identification of these hazardous compounds and formed aging products, which should be also classified as potential dangers, is of great interest.1 This contribution focused on the identification of aging products emerging in electrolytes from LIB cells using sulfur-based electrolyte additives of different chemical classes such as sulfates, sulfites, sultones and sulfonates. Ion chromatography and gas chromatography hyphenated to high-resolution acurrate mass spectrometry (HRAM-MS) were used to obtain information on the formation of ionic and volatile compounds after electrochemical operation and to elucidate the corresponding structures. Thus, specific ionic and volatile aging marker molecules could be defined for target-analysis of the original electrolyte additive in LIB material, despite a potentially complete consumption during cycling. This allows improved reverse-engineering of LIB post-mortem analysis and risk assessment. Furthermore, mechanistic conclusions on the decomposition of the investigated electrolyte additives could be drawn, extending literature reports based on X-Ray methods with respect to species information and ultimately contribute to a better understanding of the interphase constitutions.[1] C. Peschel, S. van Wickeren, A. Bloch, C. Lechtenfeld, M. Winter, S. Nowak, Energy Technol. 2022, xxx.[2] S. S. Zhang, J. Power Sources 2006, 162, 1379–1394.[3] J. Henschel, J. M. Dressler, M. Winter, S. Nowak, Chem. Mater. 2019 31 (24), 9970-9976.[4] B. Tong, Z. Song, H. Wan, W. Feng, M. Armand, J. Liu, H. Zhang, Z. Zhou, InfoMat 2021, 3, 1364–1392.

Models Genesis.

Transfer learning from natural image to medical image has been established as one of the most practical paradigms in deep learning for medical image analysis. To fit this paradigm, however, 3D imaging tasks in the most prominent imaging modalities (e.g., CT and MRI) have to be reformulated and solved in 2D, losing rich 3D anatomical information, thereby inevitably compromising its performance. To overcome this limitation, we have built a set of models, called Generic Autodidactic Models, nicknamed Models Genesis, because they are created ex nihilo (with no manual labeling), self-taught (learnt by self-supervision), and generic (served as source models for generating application-specific target models). Our extensive experiments demonstrate that our Models Genesis significantly outperform learning from scratch and existing pre-trained 3D models in all five target 3D applications covering both segmentation and classification. More importantly, learning a model from scratch simply in 3D may not necessarily yield performance better than transfer learning from ImageNet in 2D, but our Models Genesis consistently top any 2D/2.5D approaches including fine-tuning the models pre-trained from ImageNet as well as fine-tuning the 2D versions of our Models Genesis, confirming the importance of 3D anatomical information and significance of Models Genesis for 3D medical imaging. This performance is attributed to our unified self-supervised learning framework, built on a simple yet powerful observation: the sophisticated and recurrent anatomy in medical images can serve as strong yet free supervision signals for deep models to learn common anatomical representation automatically via self-supervision. As open science, all codes and pre-trained Models Genesis are available at https://github.com/MrGiovanni/ModelsGenesis.

Training deep neural networks for wireless sensor networks using loosely and weakly labeled images

Although deep learning has achieved remarkable successes over the past years, few reports have been published about applying deep neural networks to Wireless Sensor Networks (WSNs) for image targets recognition where data, energy, computation resources are limited. In this work, a Cost-Effective Domain Generalization (CEDG) algorithm has been proposed to train an efficient network with minimum labor requirements. CEDG transfers networks from a publicly available source domain to an application-specific target domain through an automatically allocated synthetic domain. The target domain is isolated from parameters tuning and used for model selection and testing only. The target domain is significantly different from the source domain because it has new target categories and is consisted of low-quality images that are out of focus, low in resolution, low in illumination, low in photographing angle. The trained network has about 7 M (ResNet-20 is about 41 M) multiplications per prediction that is small enough to allow a digital signal processor chip to do real-time recognitions in our WSN. The category-level averaged error on the unseen and unbalanced target domain has been decreased by 41.12%.

Target coverage computation protocols in wireless sensor networks: a comprehensive review

Wireless sensor network comprises tiny sensor nodes having the ability of sensing to surveil an area of interest. The amount of area sensed by a sensor node is known as coverage. Coverage is a multi-dimensional quality of service (QoS) metric in terms of diversity in its computation algorithms according to the real-time applications. Different applications require a distinctive level of coverage. But every coverage computation algorithm has one common objective, i.e. to maximize the network’s lifespan and coverage quality with minimum number of sensor nodes. Several algorithms have derived optimal or close-to-optimal solutions based on similar or distinct parameters. This paper presents a concise review of various target coverage protocols for both static and mobile targets. Total target coverage provided by the deployed sensor nodes is a measure of sensing quality. Target coverage is bifurcated as static coverage and dynamic coverage. This paper highlights and describes various application specific target coverage protocols and their QoS metrics and parameters. Each protocol has solved multi-dimensional coverage issues and challenges and proposed optimum solutions. The comprehensive review of the target coverage protocol helps in understanding about various research challenges and its existed solutions in target coverage, detection and tracking in a better way.

CASAIR: Content and Shape-Aware Image Retargeting and Its Applications.

This paper proposes a novel image-retargeting algorithm that can retarget images to a large family of non-rectangular shapes. Specifically, we study image retargeting from a broader perspective that includes the content as well as the shape of an image, and the proposed content and shape-aware image-retargeting (CASAIR) algorithm is driven by the dual objectives of image content preservation and image domain transformation, with the latter defined by an application-specific target shape. The algorithm is based on the idea of seam segment carving that successively removes low-cost seam segments from the image to simultaneously achieve the two objectives, with the selection of seam segments determined by a cost function incorporating inputs from image content and target shape. To provide a complete characterization of shapes that can be obtained using CASAIR, we introduce the notion of bhv-convex shapes, and we show that bhv-convex shapes are precisely the family of shapes that can be retargeted to by CASAIR. The proposed algorithm is simple in both its design and implementation, and in practice, it offers an efficient and effective retargeting platform that provides its users with considerable flexibility in choosing target shapes. To demonstrate the potential of CASAIR for broadening the application scope of image retargeting, this paper also proposes a smart camera-projector system that incorporates CASAIR. In the context of ubiquitous display, CASAIR equips the camera-projector system with the capability of retargeting images online in order to maximize the quality and fidelity of the displayed images whenever the situation demands.