Goto

Collaborating Authors

 Bend


Thin-Shell Object Manipulations With Differentiable Physics Simulations

arXiv.org Artificial Intelligence

In this work, we aim to teach robots to manipulate various thin-shell materials. Prior works studying thin-shell object manipulation mostly rely on heuristic policies or learn policies from real-world video demonstrations, and only focus on limited material types and tasks (e.g., cloth unfolding). However, these approaches face significant challenges when extended to a wider variety of thinshell materials and a diverse range of tasks. On the other hand, while virtual simulations are shown to be effective in diverse robot skill learning and evaluation, prior thin-shell simulation environments only support a subset of thin-shell materials, which also limits their supported range of tasks. To fill in this gap, we introduce ThinShellLab - a fully differentiable simulation platform tailored for robotic interactions with diverse thin-shell materials possessing varying material properties, enabling flexible thin-shell manipulation skill learning and evaluation. Building on top of our developed simulation engine, we design a diverse set of manipulation tasks centered around different thin-shell objects. Our experiments suggest that manipulating thin-shell objects presents several unique challenges: 1) thin-shell manipulation relies heavily on frictional forces due to the objects' co-dimensional nature, 2) the materials being manipulated are highly sensitive to minimal variations in interaction actions, and 3) the constant and frequent alteration in contact pairs makes trajectory optimization methods susceptible to local optima, and neither standard reinforcement learning algorithms nor trajectory optimization methods (either gradient-based or gradient-free) are able to solve the tasks alone. To overcome these challenges, we present an optimization scheme that couples sampling-based trajectory optimization and gradient-based optimization, boosting both learning efficiency and converged performance across various proposed tasks. By tuning simulation parameters with a minimal set of real-world data, we demonstrate successful deployment of the learned skills to real-robot settings. Manipulating thin-shell materials is complicated due to a diverse range of sophisticated activities involved in the manipulation process. For example, to lift an object using a sheet of paper, we would instinctively create a slight bend or curve in the paper before initiating the lift (Figure 1 (a)). Human beings intuitively learn such thin-shell manipulation skills, such as folding a paper to make a crease, drawing out a piece of sheet under a bottle, and even complicated card tricks. Compared with manipulating rigid bodies or volumetric materials, manipulating thin-shell materials poses several unique challenges. First, the physical forms of such materials are difficult to handle. For example, picking up a flat sheet is intrinsically difficult due to its close-to-zero thickness, preventing any effective grasping from the top.


Passive Shape Locking for Multi-Bend Growing Inflated Beam Robots

arXiv.org Artificial Intelligence

Shape change enables new capabilities for robots. One class of robots capable of dramatic shape change is soft growing "vine" robots. These robots usually feature global actuation methods for bending that limit them to simple, constant-curvature shapes. Achieving more complex "multi-bend" configurations has also been explored but requires choosing the desired configuration ahead of time, exploiting contact with the environment to maintain previous bends, or using pneumatic actuation for shape locking. In this paper, we present a novel design that enables passive, on-demand shape locking. Our design leverages a passive tip mount to apply hook-and-loop fasteners that hold bends without any pneumatic or electrical input. We characterize the robot's kinematics and ability to hold locked bends. We also experimentally evaluate the effect of hook-and-loop fasteners on beam and joint stiffness. Finally, we demonstrate our proof-of-concept prototype in 2D. Our passive shape locking design is a step towards easily reconfigurable robots that are lightweight, low-cost, and low-power.


The Human Side of AI: Predicting Spine Surgery Outcomes

#artificialintelligence

Ever since Corey Walker, MD, became a spine surgeon, the traditional measure of success focused on how well a patient was able to walk, bend or move after spine surgery. Now, with the help of artificial intelligence, Walker is measuring success differently. "The unique thing we're doing with this project is really focusing in on the pain medication part of it, because opioid addiction continues to be a challenge, and we are looking for ways to improve pain management after surgery," Walker said. Walker's team, in collaboration with the Cedars-Sinai Department of Computational Biomedicine, is using artificial intelligence and machine learning to predict which patients are most likely to successfully manage their pain post-surgery, and which patients might need additional assistance. "This project uses artificial intelligence algorithms to analyze millions of data points and predict which patients may need additional help with pain management after surgery," said Jason Moore, PhD, chair of the Department of Computational Biomedicine and acting professor of Medicine.


๐Ÿ‡บ๐Ÿ‡ธ Remote Machine learning job: Senior AI Programmer at PlayStation (Bend, Oregon, United States)

#artificialintelligence

Senior AI Programmer at PlayStation United States โ€บ Oregon โ€บ Bend (Posted May 26 2022) Do they allow remote work? Remote work is possible, see the description below for more information. PlayStation isn't just the Best Place to Play -- it's also the Best Place to Work. Today, we're recognized as a global leader in entertainment producing The PlayStation family of products and services including PlayStation 5, PlayStation 4, PlayStation VR, PlayStation Plus, PlayStation Now, acclaimed PlayStation software titles from PlayStation Studios, and more. PlayStation also strives to create an inclusive environment that empowers employees and embraces diversity.


Stretchable distributed fiber-optic sensors

Science

Distributed fiber-optic sensors have been used for monitoring mechanical deformations in stiff infrastructures such as bridges, roads, and buildings, but they either are limited to measuring one variable or require complex optics to measure multiple properties. Bai et al. now demonstrate dual-core elastomeric optical fibers, one of which contains patterned dye regions. The waveguides are fabricated by molding out of commercially available elastomers and integrate a clear core and an adjacent core doped with up to three macroscale dye regions. Changes in optical paths in the two cores detect deformation and map it onto a color space. By monitoring changes in the color and intensity in both elastomer-based fibers, the researchers could distinguish bending, stretching, and localized pressing with a spatial resolution down to โˆผ1 centimeter. Science , this issue p. [848][1] Silica-based distributed fiber-optic sensor (DFOS) systems have been a powerful tool for sensing strain, pressure, vibration, acceleration, temperature, and humidity in inextensible structures. DFOS systems, however, are incompatible with the large strains associated with soft robotics and stretchable electronics. We develop a sensor composed of parallel assemblies of elastomeric lightguides that incorporate continuum or discrete chromatic patterns. By exploiting a combination of frustrated total internal reflection and absorption, stretchable DFOSs can distinguish and measure the locations, magnitudes, and modes (stretch, bend, or press) of mechanical deformation. We further demonstrate multilocation decoupling and multimodal deformation decoupling through a stretchable DFOSโ€“integrated wireless glove that can reconfigure all types of finger joint movements and external presses simultaneously, with only a single sensor in real time. [1]: /lookup/doi/10.1126/science.aba5504


Beautiful Future: How Deschutes Uses Artificial Intelligence & Machine Learning to Brew Better Beer

#artificialintelligence

Ask any brewer and they'll admit that while beer has likely been around since the dawn of civilization, we're all still learning new ways to brew it more efficiently, creatively, and quickly. But balancing the brewer's art with modern approaches to automation, measurement, and decision making requires brewers to toe a fine line. Take the personality out of the process, and you sacrifice the "craft" in craft beer. Ignore the best tools available, and you waste precious resources that could be better spent on the creative side of the brewing equation. From their outpost on the eastern edge of the Cascades in Bend, Oregon, Deschutes Brewery has tackled this problem in a forward-thinking way, embracing their brew team's passion for tech and programming. Through their operational technology team, they're using a cutting-edge approach to brewing technology aimed at saving time and money, making higher-quality beer, and in turn freeing up company resources for an aggressive innovation program.