Showing papers by "Worcester Polytechnic Institute published in 2021"

Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators

Abstract: It is widely known that neural networks (NNs) are universal approximators of continuous functions. However, a less known but powerful result is that a NN with a single hidden layer can accurately approximate any nonlinear continuous operator. This universal approximation theorem of operators is suggestive of the structure and potential of deep neural networks (DNNs) in learning continuous operators or complex systems from streams of scattered data. Here, we thus extend this theorem to DNNs. We design a new network with small generalization error, the deep operator network (DeepONet), which consists of a DNN for encoding the discrete input function space (branch net) and another DNN for encoding the domain of the output functions (trunk net). We demonstrate that DeepONet can learn various explicit operators, such as integrals and fractional Laplacians, as well as implicit operators that represent deterministic and stochastic differential equations. We study different formulations of the input function space and its effect on the generalization error for 16 different diverse applications. Neural networks are known as universal approximators of continuous functions, but they can also approximate any mathematical operator (mapping a function to another function), which is an important capability for complex systems such as robotics control. A new deep neural network called DeepONet can lean various mathematical operators with small generalization error.

Age of Information: An Introduction and Survey

Abstract: We summarize recent contributions in the broad area of age of information (AoI). In particular, we describe the current state of the art in the design and optimization of low-latency cyberphysical systems and applications in which sources send time-stamped status updates to interested recipients. These applications desire status updates at the recipients to be as timely as possible; however, this is typically constrained by limited system resources. We describe AoI timeliness metrics and present general methods of AoI evaluation analysis that are applicable to a wide variety of sources and systems. Starting from elementary single-server queues, we apply these AoI methods to a range of increasingly complex systems, including energy harvesting sensors transmitting over noisy channels, parallel server systems, queueing networks, and various single-hop and multi-hop wireless networks. We also explore how update age is related to MMSE methods of sampling, estimation and control of stochastic processes. The paper concludes with a review of efforts to employ age optimization in cyberphysical applications.

A review of direct air capture (DAC): scaling up commercial technologies and innovating for the future

Abstract: Direct air capture (DAC) can provide an impactful, engineered approach to combat climate change by removing carbon dioxide (CO2) from the air. However, to meet climate goals, DAC needs to be scaled at a rapid rate. Current DAC approaches use engineered contactors filled with chemicals to repeatedly capture CO2 from the air and release high purity CO2 that can be stored or otherwise used. This review article focuses on two distinctive, commercial DAC processes to bind with CO2: solid sorbents and liquid solvents. We discuss the properties of solvents and sorbents, including mass transfer, heat transfer and chemical kinetics, as well as how these properties influence the design and cost of the DAC process. Further, we provide a novel overview of the considerations for deploying these DAC technologies, including concepts for learning-by-doing that may drive down costs and material requirements for scaling up DAC technologies.

Whole-genome doubling confers unique genetic vulnerabilities on tumour cells.

Abstract: Whole-genome doubling (WGD) is common in human cancers, occurring early in tumorigenesis and generating genetically unstable tetraploid cells that fuel tumour development1,2. Cells that undergo WGD (WGD+ cells) must adapt to accommodate their abnormal tetraploid state; however, the nature of these adaptations, and whether they confer vulnerabilities that can be exploited therapeutically, is unclear. Here, using sequencing data from roughly 10,000 primary human cancer samples and essentiality data from approximately 600 cancer cell lines, we show that WGD gives rise to common genetic traits that are accompanied by unique vulnerabilities. We reveal that WGD+ cells are more dependent than WGD- cells on signalling from the spindle-assembly checkpoint, DNA-replication factors and proteasome function. We also identify KIF18A, which encodes a mitotic kinesin protein, as being specifically required for the viability of WGD+ cells. Although KIF18A is largely dispensable for accurate chromosome segregation during mitosis in WGD- cells, its loss induces notable mitotic errors in WGD+ cells, ultimately impairing cell viability. Collectively, our results suggest new strategies for specifically targeting WGD+ cancer cells while sparing the normal, non-transformed WGD- cells that comprise human tissue.

Integrating sustainability and resilience in the supply chain: A systematic literature review and a research agenda

Abstract: Sustainability has emerged as an important industrial strategic outlook expanding beyond organizational boundaries to include the supply chain. Simultaneously, the industry has also been faced with supply chain resilience concerns. Research on the intersection of supply chain sustainability and resilience is nascent and is a consequence of their observed mutual influences. However, confusion about concepts, implementation methods, and measurements of sustainable and resilient supply chains remains. This study completes a systematic literature review that critically examines several major observations and directions. We find the concept of sustainable supply chains is more established, and general agreement on its theoretical foundations exists. Supply chain resilience is relatively less mature. The nexus and relationships between the two topics are often incoherent: there is confusion on sustainable and resilient supply chains establishment; there is no clarity on what practices could jointly advance both areas. A major conflict exists since sustainability generally focuses on efficiency, while resilience seeks effectiveness. We recommend studies to analyze implementation relationships and impact. We also observe that performance measurement systems should be developed to assess supply chain sustainability and resilience performance taking with explicit consideration time horizons considered in these measures.

Digitalizing the Closing-of-the-Loop for Supply Chains: A Transportation and Blockchain Perspective

Abstract: The circular economy is gaining in importance globally and locally. The COVID-19 crisis, as an exceptional event, showed the limits and the fragility of supply chains, with circular economy practices as a potential solution during and post-COVID. Reverse logistics (RL) is an important dimension of the circular economy which allows management of economic, social, and environmental challenges. Transportation is needed for RL to effectively operate, but research study on this topic has been relatively limited. New digitalization opportunities can enhance transportation and RL, and therefore further enhance the circular economy. This paper proposes to review practical research and concerns at the nexus of transportation, RL, and blockchain as a digitalizing technology. The potential benefits of blockchain technology through example use cases on various aspects of RL and transportation activities are presented. This integration and applications are evaluated using various capability facets of blockchain technology, particularly as an immutable and reliable ledger, a tracking service, a smart contract utility, as marketplace support, and as tokenization and incentivization. We also briefly introduce the physical internet concept within this context. The physical internet paradigm proposed last decade, promises to also disrupt the blockchain, transportation, and RL nexus. We include potential research directions and managerial implications across the blockchain, transportation, and RL nexus.

Experimental and Theoretical Advances in MXene-Based Gas Sensors.

Abstract: MXenes, two-dimensional (2D) transition metal carbides and nitrides, have been arousing interest lately in the field of gas sensing thanks to their remarkable features such as graphene-like morphology, metal-comparable conductivity, large surface-to-volume ratio, mechanical flexibility, and great hydrophilic surface functionalities. With tunable etching and synthesis methods, the morphology of the MXenes, the interlayer structures, and functional group ratios on their surfaces were effectively harnessed, enhancing the efficiency of MXene-based gas-sensing devices. MXenes also efficiently form nanohybrids with other nanomaterials, as a practical approach to revamp the sensing performance of the MXene sensors. This Mini-Review summarizes the recent experimental and theoretical reports on the gas-sensing applications of MXenes and their hybrids. It also discusses the challenges and provides probable solutions that can accentuate the future perspective of MXenes in gas sensors.

Algorithmic bias: review, synthesis, and future research directions

Abstract: As firms are moving towards data-driven decision making, they are facing an emerging problem, namely, algorithmic bias. Accordingly, algorithmic systems can yield socially-biased outcomes, thereby ...

Four Decades of Public Participation in Risk Decision Making.

Abstract: Over the past four decades, the promise of public participation to improve decisions, obtain legitimacy, and build capacity for risk decision making and management has had a mixed record. In this article, we offer a narrative of how public participation has evolved in the United States and we examine prospects for its future. We trace three forces that have had significant impact on practice: an emergent emphasis on democratic deliberation, a transition from dichotomous thinking about science versus politics to an integrated perspective, and the recognition that different parties to the decision-making process bring valid epistemological contributions. The promise of public participation in risk decision making is challenged by loss of trust in institutions and individuals and by broad socio-political dynamics that are weakening democratic values and processes. These include the scarcity of attitudes and aptitudes supportive of public participation among both individuals and institutions; an anti-democratic political atmosphere that promotes disrespect; pursuit of private interests over the common good; failure to appreciate the limitations of dialogue and learning; underutilization of existing knowledge; and insufficient knowledge of how context matters. We end by offering several suggestions for focusing further research and improving practice.

Ranking and Rating Bicycle Helmet Safety Performance in Oblique Impacts Using Eight Different Brain Injury Models.

Abstract: Bicycle helmets are shown to offer protection against head injuries. Rating methods and test standards are used to evaluate different helmet designs and safety performance. Both strain-based injury criteria obtained from finite element brain injury models and metrics derived from global kinematic responses can be used to evaluate helmet safety performance. Little is known about how different injury models or injury metrics would rank and rate different helmets. The objective of this study was to determine how eight brain models and eight metrics based on global kinematics rank and rate a large number of bicycle helmets (n=17) subjected to oblique impacts. The results showed that the ranking and rating are influenced by the choice of model and metric. Kendall's tau varied between 0.50 and 0.95 when the ranking was based on maximum principal strain from brain models. One specific helmet was rated as 2-star when using one brain model but as 4-star by another model. This could cause confusion for consumers rather than inform them of the relative safety performance of a helmet. Therefore, we suggest that the biomechanics community should create a norm or recommendation for future ranking and rating methods.

Using large-scale experiments and machine learning to discover theories of human decision-making

Abstract: Predicting and understanding how people make decisions has been a long-standing goal in many fields, with quantitative models of human decision-making informing research in both the social sciences and engineering. We show how progress toward this goal can be accelerated by using large datasets to power machine-learning algorithms that are constrained to produce interpretable psychological theories. Conducting the largest experiment on risky choice to date and analyzing the results using gradient-based optimization of differentiable decision theories implemented through artificial neural networks, we were able to recapitulate historical discoveries, establish that there is room to improve on existing theories, and discover a new, more accurate model of human decision-making in a form that preserves the insights from centuries of research.

From Sustainable Global Value Chains to Circular Economy: Different Silos, Different Perspectives but Many Opportunities to Build Bridges

Abstract: A growing interest in the circular economy concept has pushed the discourse in various management-related disciplines beyond established boundaries, with calls to better address how such a model may be developed in a world of global value chains. Still, the conventional linear economy model continues to dominate business, society, and research. While the concept of better connecting physical output and input flows at multiple production or consumption levels is becoming more accepted, it remains unclear how to make this happen while ensuring that sustainability targets are met or exceeded. Multiple scientific communities contribute different perspectives to this discourse, with promising opportunities for research. Circular economy and sustainability from business and economics perspectives are multifaceted. The existing body of knowledge needs to be advanced to assist private individuals, business managers, investors, or policymakers in making informed decisions. In this article for the inaugural issue, we provide a snapshot of the discourses among those who have studied the circular economy and its related topics. We outline conceptual inroads and potential research questions to encourage further circular economy and sustainability research and discourse from business or economics perspectives as well as from the broader transdisciplinary angle. We propose three research pathways: (1) connecting output with input needs in a global circular economy; (2) beyond today’s business logic for a global circular economy; and (3) inclusion of the Global South in North-dominated circular economies. For each, we propose concepts, theories, or methodological approaches and offer various perspectives from the micro, macro, and meso levels.

Evolution and Impact of Wi-Fi Technology and Applications: A Historical Perspective

Abstract: The IEEE 802.11 standard for wireless local area networking (WLAN), commercially known as Wi-Fi, has become a necessity in our day-to-day life. Over a billion Wi-Fi access points connect close to hundred billion of IoT devices, smart phones, tablets, laptops, desktops, smart TVs, video cameras, monitors, printers, and other consumer devices to the Internet to enable millions of applications to reach everyone, everywhere. The evolution of Wi-Fi technology also resulted in the first commercial piloting of spread spectrum, high speed optical communications, OFDM, MIMO and mmWave pulse transmission technologies, which then became more broadly adopted by cellular phone and wireless sensor networking industries. The popularity and widespread Wi-Fi deployment in indoor areas further motivated innovation in opportunistic cyberspace applications that exploit the ubiquitous Wi-Fi signals. The RF signal radiated from Wi-Fi access points creates an “RF cloud” accessible to any Wi-Fi equipped device hosting or supporting these opportunistic applications. Wi-Fi positioning and location intelligence were the first popular opportunistic applications of Wi-Fi’s RF cloud. Today, researchers are investigating opportunistic applications of Wi-Fi signals for gesture and motion detection as well as authentication and security. This paper provides a holistic overview of the evolution of Wi-Fi technology and its applications as the authors experienced it in the last few decades.

Hardware Trust and Assurance through Reverse Engineering: A Tutorial and Outlook from Image Analysis and Machine Learning Perspectives

Abstract: In the context of hardware trust and assurance, reverse engineering has been often considered as an illegal action. Generally speaking, reverse engineering aims to retrieve information from a produ...

An Origami Continuum Robot Capable of Precise Motion Through Torsionally Stiff Body and Smooth Inverse Kinematics

Abstract: Continuum robot arms, with their hyper-redundant continuously deformable bodies, show great promise in applications deemed impossible for traditional rigid robot arms with discrete links and joints, such as navigating tight corners without getting stuck. However, existing continuum robots suffer from excessive twisting when subjected to offset loading, even resulting from their own body weight, which reduces their dexterity and precision. In this work, we present a continuum manipulator that is capable of providing passive torsional stiffness through an origami-inspired modular design, remedying the non-controllable twist typically present in continuum robots. Our proposed origami continuum module is ∼73 times stronger in torsion compared with similar-size continuum modules made out of silicone rubber, while being 50% lighter, and capable of 125% change in length. Building on these physical capabilities, we present an optimization-based method to solve for the inverse kinematics of our multi-segment origami continuum manipulator that ensures smooth motion to follow desired end-effector paths, minimizing vibrations of the long and slender body. Further, taking advantage of the length-change capabilities of our origami manipulator, we devise and evaluate grow-to-shape algorithms to plan for full-body robot insertion motions that follow tortuous paths. Lastly, we showcase various applications of our proposed continuum robot for pick-and-place, inspection/exploration, and robotic art. Our study presents a highly capable continuum robot for safe manipulation and structure inspection applications, with potential for real-world deployment.