Search for: All records

The field of human-robot interaction has been rapidly expanding but an ever-present obstacle facing this field is developing accessible, reliable, and effective forms of communication. It is often imperative to the efficacy of the robot and the overall human-robot interaction that a robot be capable of expressing information about itself to humans in the environment. Amidst the evolving approaches to this obstacle is the use of light as a communication modality. Light-based communication effectively captures attention, can be seen at a distance, and is commonly utilized in our daily lives. Our team explored the ways light-based signals on robots are being used to improve human understanding of robot operating state. In other words, we sought to determine how light-based signals are being used to help individuals identify the conditions (e.g., capabilities, goals, needs) that comprise and dictate a robot’s current functionality. We identified four operating states (e.g., “Blocked”, “Error”, “Seeking Interaction”, “Not Seeking Interaction”) in which light is utilized to increase individuals’ understanding of the robot’s operations. These operating states are expressed through manipulation of three visual dimensions of the onboard lighting features of robots (e.g., color, pattern of lighting, frequency of pattern). In our work, we outline how these dimensions vary across operating states and the effect they have on human understanding. We also provide potential explanations for the importance of each dimension. Additionally, we discuss the main shortcomings of this technology. The first is the overlapping use of combinations of dimensions across operating states. The remainder relate to the difficulties of leveraging color to convey information. Finally, we provide considerations on how this technology might be improved going into the future through the standardization of light-based signals and increasing the amount of information provided within interactions between agents. 

Abstract Due to the global pandemic in 2020, rehabilitation services had to adapt to virtual care to meet the needs of patients with a diversity of upper extremity disorders and injuries, from simple fingertip injuries to replanted extremities, while mitigating the spread of COVID-19. For patients who require occupational therapy in their rehabilitation, teleoperated robotic care can fulfill this need. Moreover, some clinical studies show that incorporating the active intention of patients into rehabilitation training enhances the effectiveness of therapies. This research aims to develop a bilateral impedance controller with a Franka Emika 7DOF (degree-of-freedom) articulated robot using matlab, facilitating upper limb occupational therapy. Through this setup, the Franka robot will lead patients through various therapeutic exercises based on their motion intention, such as a pick-and-place exercise. The exercises' efficacy is assessed through evaluations of torque, muscle strength, and jerking performance. During patients' rehabilitation exercises with the Franka robot, we anticipate observing decreased muscle activity alongside a smooth, assist-as-needed (AAN) trajectory. This method will allow a patient to perform complex movements essential for daily activities quicker than traditional occupational therapy. 

NMR metabolomics was applied toM. mercenariaexposed to a series of sublethal doses of Brevetoxins for the targeted metabolic pathway and early marker discovery. 

Abstract While the positive relationship between plant diversity and ecosystem functioning is frequently observed and often attributed to direct plant–plant interactions, it remains unclear whether and how the effects of plant diversity endure through soil legacy effects, particularly at the level of genotypic diversity. We manipulated the genotypic diversity ofScirpus mariqueterand tested its soil legacy effects on a conspecific phytometer under low‐ and high‐water availability conditions. We found that genotypic diversity enhanced phytometer productivity through soil legacies, with stronger effects under low‐water availability conditions, improving its resistance to water stress. Moreover, this effect was attributed to the association between asexual and sexual reproductive strategies by increasing ramet number to ensure plant survival under low‐water availability and promoting sexual reproduction to escape stress. The observed diversity effects were primarily associated with increased levels of microbial biomass in soils trained by populations with diverse genotypes. Our findings highlight the importance of plant genotypic diversity in modulating ecosystem functioning through soil legacies and call for management measures that promote genetic diversity to make ecosystems sustainable in the face of climate change. 

An ensemble data-learning approach based on proper orthogonal decomposition (POD) and Galerkin projection (EnPOD-GP) is proposed for thermal simulations of multi-core CPUs to improve training efficiency and the model accuracy for a previously developed global POD-GP method (GPOD-GP). GPOD-GP generates one set of basis functions (or POD modes) to account for thermal behavior in response to variations in dynamic power maps (PMs) in the entire chip, which is computationally intensive to cover possible variations of all power sources. EnPOD-GP however acquires multiple sets of POD modes to significantly improve training efficiency and effectiveness, and its simulation accuracy is independent of any dynamic PM. Compared to finite element simulation, both GPOD-GP and EnPOD-GP offer a computational speedup over 3 orders of magnitude. For a processor with a small number of cores, GPOD-GP provides a more efficient approach. When high accuracy is desired and/or a processor with more cores is involved, EnPOD-GP is more preferable in terms of training effort and simulation accuracy and efficiency. Additionally, the error resulting from EnPOD-GP can be precisely predicted for any random spatiotemporal power excitation. 

The classical proper orthogonal decomposition (POD) with the Galerkin projection (GP) has been revised for chip-level thermal simulation of microprocessors with a large number of cores. An ensemble POD-GP methodology (EnPODGP) is introduced to significantly improve the training effectiveness and prediction accuracy by dividing a large number of heat sources into heat source blocks (HSBs) each of which may contains one or a very small number of heat sources. Although very accurate, efficient and robust to any power map, EnPOD-GP suffers from intensive training for microprocessors with an enormous number of cores. A local-domain EnPOD-GP model (LEnPOD-GP) is thus proposed to further minimize the training burden. LEnPOD-GP utilizes the concepts of local domain truncation and generic building blocks to reduce the massive training data. LEnPOD-GP has been demonstrated on thermal simulation of NVIDIA Tesla Volta™ GV100, a GPU with more than 13,000 cores including FP32, FP64, INT32, and Tensor Cores. Due to the domain truncation for LEnPOD-GP, the least square error (LSE) is degraded but is still as small as 1.6% over the entire space and below 1.4% in the device layer when using 4 modes per HSB. When only the maximum temperature of the entire GPU is of interest, LEnPOD-GP offers a computing speed 1.1 million times faster than the FEM with a maximum error near 1.2oC over the entire simulation time.