Search for: All records

Energy-efficient image acquisition on the edge is crucial for enabling remote sensing applications where the sensor node has weak compute capabilities and must transmit data to a remote server/cloud for processing. To reduce the edge energy consumption, this paper proposes a sensor-algorithm co-designed system called SNAPPIX, which compresses raw pixels in the analog domain inside the sensor. We use coded exposure (CE) as the in-sensor compression strategy as it offers the flexibility to sample, i.e., selectively expose pixels, both spatially and temporally. SNAPPIX has three contributions. First, we propose a task-agnostic strategy to learn the sampling/exposure pattern based on the classic theory of efficient coding. Second, we co- design the downstream vision model with the exposure pattern to address the pixel-level non-uniformity unique to CE-compressed images. Finally, we propose lightweight augmentations to the image sensor hardware to support our in-sensor CE compres- sion. Evaluating on action recognition and video reconstruction, SNAPPIX outperforms state-of-the-art video-based methods at the same speed while reducing the energy by up to 15.4×. We have open-sourced the code at: https://github.com/horizon- research/SnapPix. 

A precise dynamical characterization of quantum impurity models with multiple interacting orbitals is challenging. In quantum Monte Carlo methods, this is embodied by sign problems. A dynamical sign problem makes it exponentially difficult to simulate long times. A multi-orbital sign problem generally results in a prohibitive computational cost for systems with multiple impurity degrees of freedom even in static equilibrium calculations. Here, we present a numerically exact inchworm method that simultaneously alleviates both sign problems, enabling simulation of multi-orbital systems directly in the equilibrium or nonequilibrium steady-state. The method combines ideas from the recently developed steady-state inchworm Monte Carlo framework [Erpenbeck et al., Phys. Rev. Lett. 130, 186301 (2023)] with other ideas from the equilibrium multi-orbital inchworm algorithm [Eidelstein et al., Phys. Rev. Lett. 124, 206405 (2020)]. We verify our method by comparison with analytical limits and numerical results from previous methods. 

The extremely large slip that occurred on the shallow portion of the Japan Trench subduction zone during the 2011 Mw 9.1 Tohoku-oki earthquake directly contributed to the devastating tsunami that inundated the Pacific coast of Japan. International Ocean Discovery Program (IODP) Expedition 405 (Tracking Tsunamigenic Slip Across the Japan Trench) aimed to investigate the conditions and processes that facilitated the extremely shallow slip on the subduction interface during the 2011 Tohoku-oki earthquake to improve understanding of the factors that allow slip to the trench on subduction zones. Expedition 405 implemented a combined logging, coring, and observatory operational plan at two sites: Site C0026, ~8 km seaward of the Japan Trench, to characterize the input sediments to the subduction zone and Site C0019, ~6 km landward of the trench, where the plate boundary fault zone is present at ~825 meters below seafloor (mbsf). At Site C0026, the input section was logged to ~430 mbsf with a logging-while-drilling (LWD) assembly that characterized the succession of sediments and rocks from the seafloor to the basaltic rocks of the oceanic crust. Cores recovered from four holes as deep as 290 mbsf contain a sequence of hemipelagic and pelagic sediments that will be input into the shallow subduction system and therefore control both the localization of the plate boundary fault zone and the slip behavior of the plate boundary. Site C0019 was previously drilled in 2012 during Integrated Ocean Drilling Program Expedition 343 (Japan Trench Fast Drilling Project [JFAST]), and revisiting this site allowed temporal variations in the frontal prism and plate boundary fault zone to be evaluated. The LWD data to ~980 mbsf characterized the frontal prism, plate boundary fault zone, and lower plate to the basaltic volcanic rocks. Cores were recovered from multiple holes that contain a variety of muds from the frontal prism and the plate boundary fault zone, as well as lower plate materials. Comparison with the sediments from Site C0026 provides a basis to interpret the tectonic and sedimentological processes operating in the dynamic environment of the frontal prism. Cores from the plate boundary fault zone provide a unique window into the structural complexity of an active plate boundary fault that is known to host large seismic slip. Two borehole observatories were installed at Site C0019 that contain temperature sensors deployed to take measurements over a period of years and reveal the hydrogeologic structure of the shallow subduction system. These hugely successful drilling operations, combined with postexpedition work to measure the mechanical, frictional, paleomagnetic, and hydrogeologic properties of the core samples and to constrain the history of past seismic slip at Site C0019, will provide an unprecedented opportunity to advance our understanding of shallow subduction systems. Outreach during the expedition leveraged and elevated the success of the operations by sharing the outcomes with a variety of domestic and international audiences, including scientists, students, educators, stakeholders, and the general public. Thanks to the efforts of a large group of onboard outreach officers and their onshore support, activities included ship-to-shore broadcast events; interviews with science party members and crew; the publication of videos, blogs, magazine articles, and social media posts; and development of formalized classroom lesson plans and materials. 

Frontal polymerization (FP) is a promising alternative manufacturing method for thermoset-based fiber-reinforced polymer composites (FRP) in comparison with the traditional autoclave/oven-curing method, due to its rapid curing process, low energy consumption, and low cost. Optimizing the weight contents of initiators relative to the resin’s mass is needed to adjust the mechanical properties of FRPs in industrial applications. This study investigates the effect of varying the photo-initiator (PI) weight content on tensile properties and the frontal polymerization characteristics, including the front velocity, front temperature, and degree of cure, in the FP process of the epoxy resin. Specifically, a dual-initiator system, including PI and thermal-initiator (TI), is used to initiate the polymerization process by ultraviolent (UV) light. The weight content of the TI is fixed at 1 w%, and the relative PI concentration is varied from 0.2 w% to 0.5 wt%. Results show that increasing the PI amount from 0.2 wt% to 0.3 wt% significantly improves the front velocity and the degree of cure by about two times. Increasing the PI content from 0.3 wt% to 0.4 wt% results in 15% and 26% higher degree of cure and front velocity, respectively. Moreover, due to the different front velocity in the top and bottom regions of the specimen, the specimens with 0.4 wt% PI exhibited a curved shape. The specimen with 0.5 wt% PI is thermally degraded and foamed. By comparing tensile properties, it is found that increasing the PI concentration from 0.2 wt% to 0.3 wt% improves the tensile strength and Young’s modulus by 3.91% and 7%, respectively, while the tensile strength and the Young’s modulus of frontal polymerized specimens are on average 8% and 14% higher than traditionally oven-cured ones, respectively.