Catalogue Search | MBRL
Are you sure you want to remove the book from the shelf?
{{itemTitle}}
110
result(s) for
"Sarma, Sanjay"
Sort by:
Selective learning for sensing using shift-invariant spectrally stable undersampled networks
The amount of data collected for sensing tasks in scientific computing is based on the Shannon-Nyquist sampling theorem proposed in the 1940s. Sensor data generation will surpass 73 trillion GB by 2025 as we increase the high-fidelity digitization of the physical world. Skyrocketing data infrastructure costs and time to maintain and compute on all this data are increasingly common. To address this, we introduce a selective learning approach, where the amount of data collected is problem dependent. We develop novel shift-invariant and spectrally stable neural networks to solve real-time sensing problems formulated as classification or regression problems. We demonstrate that (i) less data can be collected while preserving information, and (ii) test accuracy improves with data augmentation (size of training data), rather than by collecting more than a certain fraction of raw data, unlike information theoretic approaches. While sampling at Nyquist rates, every data point does not have to be resolved at Nyquist and the network learns the amount of data to be collected. This has significant implications (orders of magnitude reduction) on the amount of data collected, computation, power, time, bandwidth, and latency required for several embedded applications ranging from low earth orbit economy to unmanned underwater vehicles.
Journal Article
Measuring light transport properties using speckle patterns as structured illumination
The measurement of light absorption and scattering properties of biological materials has several diagnostic and therapeutic applications. We can measure these properties for skin without contact using structured illumination and imaging. However, building simple handheld devices remains challenging due to motion artefacts and moving targets. To overcome this limitation, we project random speckle patterns instead of discrete spatial frequencies on the target. Since random patterns are spatially broadband, they capture more information per image, enabling frame-by-frame analysis. In this paper, we describe the statistics of objective speckles and demonstrate how the optical system is designed for spatially bandlimited illumination. Next, we use a diverse set of liquid tissue phantom to validate the method. We successfully demonstrate that a calibrated instrument can independently predict the two primary light transport properties of a homogeneous turbid system. This work is a starting point for analysing skin and other heterogeneous biological media in the future.
Journal Article
Mimicking Sub-Structures Self-Organization in Microtubules
Microtubules (MTs) are highly dynamic polymers distributed in the cytoplasm of a biological cell. Alpha and beta globular proteins constituting the heterodimer building blocks combine to form these tubules through polymerization, controlled by the concentration of Guanosine-triphosphate (GTPs) and other Microtubule Associated Proteins (MAPs). MTs play a crucial role in many intracellular processes, predominantly in mitosis, organelle transport and cell locomotion. Current research in this area is focused on understanding the exclusive behaviors of self-organization and their association with different MAPs through organized laboratory experiments. However, the intriguing intelligence behind these tiny machines resulting in complex self-organizing structures is mostly unexplored. In this study, we propose a novel swarm engineering framework in modeling rules for these systems, by combining the principles of design with swarm intelligence. The proposed framework was simulated on a game engine and these simulations demonstrated self-organization of rings and protofilaments in MTs. Analytics from these simulations assisted in understanding the influence of GTPs on protofilament formation. Also, results showed that the population density of GTPs rather than their bonding probabilities played a crucial role in polymerization in forming microtubule substructures.
Journal Article
Mechanical Ventilator Parameter Estimation for Lung Health through Machine Learning
Patients whose lungs are compromised due to various respiratory health concerns require mechanical ventilation for support in breathing. Different mechanical ventilation settings are selected depending on the patient’s lung condition, and the selection of these parameters depends on the observed patient response and experience of the clinicians involved. To support this decision-making process for clinicians, good prediction models are always beneficial in improving the setting accuracy, reducing treatment error, and quickly weaning patients off the ventilation support. In this study, we developed a machine learning model for estimation of the mechanical ventilation parameters for lung health. The model is based on inverse mapping of artificial neural networks with the Graded Particle Swarm Optimizer. In this new variant, we introduced grouping and hierarchy in the swarm in addition to the general rules of particle swarm optimization to further improve its prediction performance of the mechanical ventilation parameters. The machine learning model was trained and tested using clinical data from canine and feline patients at the University of Georgia College of Veterinary Medicine. Our model successfully generated a range of parameter values for the mechanical ventilation applied on test data, with the average prediction values over multiple trials close to the target values. Overall, the developed machine learning model should be able to predict the mechanical ventilation settings for various respiratory conditions for patient’s survival once the relevant data are available.
Journal Article
Session-based security enhancement of RFID systems for emerging open-loop applications
Radio frequency identification (RFID) is an important technique used for automatic identification and data capture. In recent years, low-cost RFID tags have been used in many open-loop applications beyond supply chain management, such as the tagging of the medicine, clothes, and belongings after the point of sales. At the same time, with the development of semiconductor industry, handheld terminals and mobile phones are becoming RFID-enabled. Unauthorized mobile RFID readers could be abused by the malicious hackers or curious common people. Even for authorized RFID readers, the ownership of the reader can be transferred and the owners of the authorized mobile reader may not be always reliable. The authorization and authentication of the mobile RFID readers need to take stronger security measures to address the privacy or security issues that may arise in the emerging open-loop applications. In this paper, the security demands of RFID tags in emerging open-loop applications are summarized, and two example protocols for authorization, authentication and key establishment based on symmetric cryptography are presented. The proposed protocols adopt a timed-session-based authorization scheme, and all reader-to-tag operations are authorized by a trusted third party using a newly defined class of timed sessions. The output of the tags is randomized to prevent unauthorized tracking of the RFID tags. An instance of the protocol A is implemented in 0.13-μm CMOS technology, and the functions are verified by field programmable gate array. The baseband consumes 44.0 μW under 1.08 V voltage and 1.92 MHz frequency, and it has 25,067 gate equivalents. The proposed protocols can successfully resist most security threats toward open-loop RFID systems except physical attacks. The timing and scalability of the two protocols are discussed in detail.
Journal Article
The Quest for Quality Jobs
The US lost one-third of its manufacturing jobs--5.8 million positions--between 2000 and 2010. Although the economy has strengthened significantly since then, only about 12% of these jobs have returned. Their disappearance has resulted in painful social disruption: manufacturing had been a critical route to the middle class for those with high school educations or less. A recent study by the Georgetown University Center on Education and the Workforce shows that in the period from 2007 to 2016--the Great Recession and its recovery--8.6 million jobs were filled with persons with a bachelor's degree or higher; 1.3 million jobs by those with an associate degree or some college; and in a decline, 5.5 million jobs by those with high school degrees or below. With deep societal challenges around quality-job creation, a general upskilling already under way, and the entry of new digital technologies, there is a growing consensus on the need for improved workforce education.
Journal Article
Fixing an Imperfect Labor Market Information System
The US has the most decentralized workforce management of any developed nation. Its labor market is highly individualized and localized, and thus fragmented with relatively few information connections among participants. On the supply side, younger workers face difficult education decisions, especially at ages 17 or 18, that will largely determine their work futures; yet they act with limited understanding about their options and the implications of their choices. Displaced workers are in a worse box; they lack the tools to understand job opening, education, and training options. Head AI, a Finnish company, is developing a \"microcompetencies\" system that maps regions, cities, and organizations showing in real time which skills are most in demand and where they are needed. It is working on system to map online an individual's skills, identify employment fields that fit this personal map, and suggest additional skill areas that would help the individual meet job demands in particular new fields.
Journal Article
Value in vehicles: economic assessment of automotive data
PurposeThe purpose of this paper is to develop a model to estimate the value of information generated by and stored within vehicles to help people, businesses and researchers.Design/methodology/approachThe authors provide a taxonomy for data within connected vehicles, as well as for actors that value such data. The authors create a monetary value model for different data generation scenarios from the perspective of multiple actors.FindingsActors value data differently depending on whether the information is kept within the vehicle or on peripheral devices. The model shows the US connected vehicle data market is worth between US$11.6bn and US$92.6bn.Research limitations/implicationsThis model estimates the value of vehicle data, but a lack of academic references for individual inputs makes finding reliable inputs difficult. The model performance is limited by the accuracy of the authors’ assumptions.Practical implicationsThe proposed model demonstrates that connected vehicle data has higher value than people and companies are aware of, and therefore we must secure these data and establish comprehensive rules pertaining to data ownership and stewardship.Social implicationsEstimating the value of data of vehicle data will help companies understand the importance of responsible data stewardship, as well as drive individuals to become more responsible digital citizens.Originality/valueThis is the first paper to propose a model for computing the monetary value of connected vehicle data, as well as the first paper to provide an estimate of this value.
Journal Article
