Search for: All records

To achieve human-level dexterity, robots must infer spatial awareness from multimodal sensing to reason over contact interactions. During in-hand manipulation of novel objects, such spatial awareness involves estimating the object’s pose and shape. The status quo for in-hand perception primarily uses vision and is restricted to tracking a priori known objects. Moreover, visual occlusion of objects in hand is imminent during manipulation, preventing current systems from pushing beyond tasks without occlusion. We combined vision and touch sensing on a multifingered hand to estimate an object’s pose and shape during in-hand manipulation. Our method, NeuralFeels, encodes object geometry by learning a neural field online and jointly tracks it by optimizing a pose graph problem. We studied multimodal in-hand perception in simulation and the real world, interacting with different objects via a proprioception-driven policy. Our experiments showed final reconstructionFscores of 81% and average pose drifts of 4.7 millimeters, which was further reduced to 2.3 millimeters with known object models. In addition, we observed that, under heavy visual occlusion, we could achieve improvements in tracking up to 94% compared with vision-only methods. Our results demonstrate that touch, at the very least, refines and, at the very best, disambiguates visual estimates during in-hand manipulation. We release our evaluation dataset of 70 experiments, FeelSight, as a step toward benchmarking in this domain. Our neural representation driven by multimodal sensing can serve as a perception backbone toward advancing robot dexterity. 

The use of stone hammers to produce sharp stone flakes—knapping—is thought to represent a significant stage in hominin technological evolution because it facilitated the exploitation of novel resources, including meat obtained from medium‐to‐large‐sized vertebrates. The invention of knapping may have occurred via an additive (i.e., cumulative) process that combined several innovative stages. Here, we propose that one of these stages was the hominin use of ‘naturaliths,’ which we define as naturally produced sharp stone fragments that could be used as cutting tools. Based on a review of the literature and our own research, we first suggest that the ‘typical’ view, namely that sharp‐edged stones are seldom produced by nonprimate processes, is likely incorrect. Instead, naturaliths can be, and are being, endlessly produced in a wide range of settings and thus may occur on the landscape in far greater numbers than archaeologists currently understand or acknowledge. We then explore the potential role this ‘naturalith prevalence’ may have played in the origin of hominin stone knapping. Our hypothesis suggests that the origin of knapping was not a ‘Eureka!’ moment whereby hominins first made a sharp flake by intention or by accident and then sought something to cut, but instead was an emulative process by hominins aiming to reproduce the sharp tools furnished by mother nature and already in demand. We conclude with a discussion of several corollaries our proposal prompts, and several avenues of future research that can support or question our proposal. 

We have prepared the hydrogen sulfide trimer and tetramer anions, (H 2 S) 3 − and (H 2 S) 4 − , measured their anion photoelectron spectra, and applied high-level quantum chemical calculations to interpret the results. The sharp peaks at low electron binding energies in their photoelectron spectra and their diffuse Dyson orbitals are evidence for them both being dipole-bound anions. While the dipole moments of the neutral (H 2 S) 3 and (H 2 S) 4 clusters are small, the excess electron induces structural distortions that enhance the charge-dipolar attraction and facilitate the binding of diffuse electrons. 

Abstract Because North and South America are surrounded by water, they constitute together a gigantic island whose peripheral sea level is controlled by the winds east of the island, winds along the western boundary of the island, the freshwater flux, and the meridional overturning cell. This idea has been expressed in several articles where a series of analytical models show that the Bering Strait (BS) flow is controlled by the interplay of the Southern Winds (sometimes referred to as the “Subantarctic Westerlies”), and the North Hemisphere freshwater flux. Here, the authors report a paleoceanographic analysis of proxies in the BS as well as the Southern Ocean, which clearly support the above through employment of a slowly varying time-dependent version of the coupled Sandal–Nof model. This study shows a very strong correlation between the Southern Ocean winds and the BS flow. A mid-Holocene weakening of the Southern Winds followed by the cession of freshwater fluxes from the melting Laurentide ice sheet strengthened the BS flow for several thousand years. Increasing the Southern Winds enhances the near surface, cross-equatorial flow from the Southern Ocean to the Northern Hemisphere. This cross-equatorial flow decreases the Arctic outflow into the Atlantic demonstrating a dynamic linkage between the Southern Ocean Winds and the mean flow through the BS.