PhysicalAI Robotics Manipulation in the Kitchen

  Dataset Description:

PhysicalAI-Robotics-Manipulation-Kitchen is a dataset of automatic generated motions of robots performing operations such as opening and closing cabinets, drawers, dishwashers and fridges. The dataset was generated in IsaacSim leveraging reasoning algorithms and optimization-based motion planning to find solutions to the tasks automatically [1, 3]. The dataset includes a bimanual manipulator built with… See the full description on the dataset page: https://huggingface.co/datasets/nvidia/PhysicalAI-Robotics-Manipulation-Kitchen.

About this webinar We rarely receive the research data in an appropriate form. Often data is messy. Sometimes it is incomplete. And sometimes there’s too much of it. Frequently, it has errors. This webinar targets beginners and presents a quick demonstration of using the most widespread data wrangling tool, Microsoft Excel, to sort, filter, copy, protect, transform, aggregate, summarise, and visualise research data. Webinar Topics Introduction to Microsoft Excel user interface Interpret data using sorting, filtering, and conditional formatting Summarise data using functions Analyse data using pivot tables Manipulate and visualise data Handy tips to speed up your work Licence Copyright © 2021 Intersect Australia Ltd. All rights reserved.

The reseach objective is to present a microfluidic approach to achieve the dynamic control of particle pathlines within a flow through microfluidic device. Our approach combines three key aspects: the design of a flow-through microfluidic flow cell with the ability to manipulate the streamlines of the flow, an optimization procedure to find a priori optimal particle path-lines, and a Proportion-Integral-Derivative-based (PID) feedback controller to provide real time control over the particle manipulations. The experimental raw images were recorded with a sCMOS camera (PCO) with a pixel pitch of 6.5 μm. The camera was mounted on a microscope (Nikon Eclipse Ti) with a 1x objective. The

acquisition frequency was 5 Hz corresponding to an average in-plane displacement of 4-6 pixels between two consecutive recordings. The zip file contains the raw images and the MATLAB script used to do an experiment of two particles coming close to each other by only using the hydrodynamic forcing in a Hele-Shaw cell.

The Geometry Manipulation Protocol (GMP) is a library which serializes datatypes between XML and ANSI C data structures to support CFD applications. This library currently provides a description of geometric configurations, general moving-body scenarios (prescribed and/or 6-DOF), and control surface settings.

This file can be used to manipulate the storage technologies cost data for the Starter Data Kits.

Introspective Planning: Aligning Robots’ Uncertainty with Inherent Task Ambiguity

🌐 Project Page | 📄 Paper | GitHub

  Overview

Safe Mobile Manipulation dataset was designed to evaluate Large Language Models' (LLMs) capability to reason effectively about both uncertainty and safety in mobile manipulation tasks. The dataset comprises 500 total scenarios: 100 scenarios in the test set, 200 scenarios for knowledge base construction, and 200 scenarios for conformal calibration.… See the full description on the dataset page: https://huggingface.co/datasets/kaiquliang/Safe-Mobile-Manipulation.

Data used in the analyses as well as files generated by simulations. The code that uses and/or produces these data files is found in Zenodo and/or GitHub.

Data acquired during large part co-manipulation processes. Specifically, trajectory percentage and trajectory deviation.

Notation of files (i.e. "u2_AB_500.csv"):

• User: u2 would be the second subject of the experiment.
• Path: Two options, AB (station A to station B) and BA (station B to A).
• Maximum allowed distance: Value which defined the width of the lane.

A dataset of multimodal sensing modalities (forces, torques, poses, RGBD images) for acquiring bites of food. Includes a diverse set of food items, from vegetables like spinach and broccoli, to bready items like sandwiches and pizza, to thick items like mashed potatoes and jello, to composite items like noodles and rice & beans. This dataset extends "A Dataset of Food Manipulation Strategies" ( https://doi.org/10.7910/DVN/8TTXZ7 ) to a much more diverse set of food items.

Source code for ToPS. A compressed file containing the source code for ToPS. (GZ)

Dataset and R code for reproducing figures and statistical results in 'Homo floresiensis did not manipulate hammer angle when making stone tools', by Lin et al.

This dataset presents experimental data collected from three cyborg cockroaches using non-invasive UV stimulation. The dataset includes 15 trials in total, with 5 trials for each cockroach.The UV stimulation was applied as a means to manipulate the cockroach's compound eyes, effectively directing their trajectories. The dataset provides detailed recordings of the insects' responses to UV stimulation, offering valuable insights into the effectiveness of this sensory manipulation for bio-hybrid systems. This study represents the first successful demonstration of using UV stimulation to control cyborg insect locomotion, and the data contained here can serve as a critical resource for advancing research in the field of biohybrid robotics and insect-based control mechanisms.

Many parasites can alter the behavior of their hosts in a manner that is beneficial for the parasite (parasitic manipulations). Obviously, examples of non-manipulative parasites also exist, however, their number might be underestimated because of the publication bias. Trematodes from the Diplostomidae family infecting fish eyes are often considered manipulators. However, only one species (Diplostomum pseudospathaceum) of this family has been shown to do so under controlled laboratory conditions. We experimentally studied whether another common diplostomid species (Tylodelphys clavata) manipulated the host defensive behavior using salmonids (Salvelinus malma) reared and infected in the laboratory. We tested fish activity, depth preference, and dip net avoidance (common fish defensive traits) under different light conditions. Although the experimental design was identical to those used earlier for D. pseudospathaceum, no manipulative abilities were detected inT. clavata. Infected fish did not differ from control in the expression of the defensive behavioral traits tested. Interestingly, fish activity was confounded with fish size in control but not in infected fish, however, this pattern does not seem to be a manipulation. Our results show that even closely related parasites occupying similar habitat niches can dramatically differ in their ability to manipulate host behavior.

Automated social media accounts, known as bots, have been shown to spread disinformation and manipulate online discussions. We study the behavior of retweet bots on Twitter during the first impeachment of U.S. President Donald Trump. We collect over 67.7 million impeachment related tweets from 3.6 million users, along with their 53.6 million edge follower network. We find although bots represent 1% of all users, they generate over 31% of all impeachment related tweets. We also find bots share more disinformation, but use less toxic language than other users. Among supporters of the Qanon conspiracy theory, a popular disinformation campaign, bots have a prevalence near 10%. The follower network of Qanon supporters exhibits a hierarchical structure, with bots acting as central hubs surrounded by isolated humans. We quantify bot impact using the generalized harmonic influence centrality measure. We find there are a greater number of pro-Trump bots, but on a per bot basis, anti-Trump and pro-Trump bots have similar impact, while Qanon bots have less impact. This lower impact is due to the homophily of the Qanon follower network, suggesting this disinformation is spread mostly within online echo-chambers.

Theory predicts a trade-off between sexually-selected weapons used to secure mates and post-copulatory traits used to maximize fertilization success. However, individuals that have a greater capacity to acquire resources from the environment may invest more in both pre- and post-copulatory traits, and trade-offs may not be readily apparent. Here, we manipulate the phenotype of developing individuals to examine allocation trade-offs between weapons and testes in Mictis profana (Hemiptera: Coreidae), a species where the hind legs are sexually selected weapons used in contests over access to females. We experimentally prevented males from developing weapons by inducing them to autotomize their hind legs before the final molt to adulthood. We compared trait expression in this group to males where autotomy was induced in the mid legs, which are presumably not under sexual selection to the same extent. We found males without weapons invested proportionally more in testes mass than those with their mid legs removed. Males that developed to adulthood without weapons did not differ from the mid leg removal group in other traits potentially under pre-copulatory sexual selection, other post-copulatory traits, or naturally selected traits. In addition, a sample of adult males from the same population in the wild revealed a positive correlation between investment in testes and weapons. Our study presents a critical contribution to a growing body of literature suggesting the allocation of resources to pre- and post-copulatory sexual traits is influenced by a resource allocation trade-off and that this trade-off may only be revealed with experimental manipulation.

The Incarceration Report Control System (IRCS) contains the reporters and facilities that submit inmate information. It controls the expiration dates for the prisoner agreements and reporting, as well as displaying the total incentive payments issued for each month based on the facility identification code.

This is a preregistration for a manipulation check for a survey experiment. The goal of the manipulation check is to test whether the intended manipulations work as we expect. The treatment consists out of a manipulation of the level of regulatory controllability (high vs low) over chemical risks. We developed this manipulation for three different types of regulatory controllability: 1) availability of legal sanctions, 2) availability of sufficient manpower to detect misconduct, and 3) availability of knowledge. In addition, we also test whether the strength of the manipulation matters. We do so by testing each treatment twice: one with a weak, and the other with a strong manipulation. Hereby the strong manipulation, in contrast to the weak manipulation, includes an explicit conclusion about the controllability of the regulator. Moreover, we use the data collection of this manipulation check to test the emotion scale we selected for the survey experiment.

There is a growing need to better understand the potential impacts of altered thermal regimes on biodiversity and ecosystem function as mean temperatures, and the likelihood of extreme temperatures, continue to increase. One valuable approach to identify mechanisms and pathways of thermally-driven change at the community level is through the manipulation of temperature in the field. However, where methods exist, they are often costly or unable to produce ecologically relevant changes in temperature. Here, we present a low cost, easily assembled, and readily customizable thermal manipulation system for tide pools or other small bodies of water – the Seaside Array for Understanding Thermal Effects (SAUTE) – and demonstrate its ability to effectively alter the temperature in tide pools. During our three-hour heating manipulation, heated pools reached temperatures 4°C warmer than unmanipulated pools. During the cooling manipulation, cooled pools remained on average 1.8°C cooler than control pools. The novel SAUTE system can be used to alter the temperature of tide pools in situ. Further, it could be modified to heat other environments such as freshwater vernal pools and settlement tiles in a realistic and meaningful manner, serving as a useful tool to test questions surrounding the relationship between climate warming, thermal variability, and ecological processes in natural aquatic communities.

REASSEMBLE: A Multimodal Dataset for Contact-rich Robotic Assembly and Disassembly

📋 Introduction

Robotic manipulation remains a core challenge in robotics, particularly for contact-rich tasks such as industrial assembly and disassembly. Existing datasets have significantly advanced learning in manipulation but are primarily focused on simpler tasks like object rearrangement, falling short of capturing the complexity and physical dynamics involved in assembly and disassembly. To bridge this gap, we present REASSEMBLE (Robotic assEmbly disASSEMBLy datasEt), a new dataset designed specifically for contact-rich manipulation tasks. Built around the NIST Assembly Task Board 1 benchmark, REASSEMBLE includes four actions (pick, insert, remove, and place) involving 17 objects. The dataset contains 4,551 demonstrations, of which 4,035 were successful, spanning a total of 781 minutes. Our dataset features multi-modal sensor data including event cameras, force-torque sensors, microphones, and multi-view RGB cameras. This diverse dataset supports research in areas such as learning contact-rich manipulation, task condition identification, action segmentation, and more. We believe REASSEMBLE will be a valuable resource for advancing robotic manipulation in complex, real-world scenarios.

{"Hama1": [
    [x ,y, z],
    [qx, qy, qz, qw]
 ], 
 "Hama2": [
    [x ,y, z],
    [qx, qy, qz, qw]
 ], 
 "DAVIS346": [
    [x ,y, z],
    [qx, qy, qz, qw]
 ], 
 "NIST_Board1": [
    [x ,y, z],
    [qx, qy, qz, qw]
 ]
}

[x, y, z] represent the position of the object, and [qx, qy, qz, qw] represent its orientation as a quaternion.

The HDF5 (.h5) format organizes data into two main types of structures: datasets, which hold the actual data, and groups, which act like folders that can contain datasets or other groups. In the diagram below, groups are shown as folder icons, and datasets as file icons. The main group of the file directly contains the video, audio, and event data. To save memory, video and audio are stored as encoded byte strings, while event data is stored as arrays. The robot’s proprioceptive information is kept in the robot_state group as arrays. Because different sensors record data at different rates, the arrays vary in length (signified by the N_xxx variable in the data shapes). To align the sensory data, each sensor’s timestamps are stored separately in the timestamps group. Information about action segments is stored in the segments_info group. Each segment is saved as a subgroup, named according to its order in the demonstration, and includes a start timestamp, end timestamp, a success indicator, and a natural language description of the action. Within each segment, low-level skills are organized under a low_level subgroup, following the same structure as the high-level annotations.

📁

The splits folder contains two text files which list the h5 files used for the traning and validation splits.

📌 Important Resources

The project website contains more details about the REASSEMBLE dataset. The Code for loading and visualizing the data is avaibile on our github repository.

📄 Project website: https://tuwien-asl.github.io/REASSEMBLE_page/
💻 Code: https://github.com/TUWIEN-ASL/REASSEMBLE

MethodsThe objective of this project was to determine the capability of a federated analysis approach using DataSHIELD to maintain the level of results of a classical centralized analysis in a real-world setting. This research was carried out on an anonymous synthetic longitudinal real-world oncology cohort randomly splitted in three local databases, mimicking three healthcare organizations, stored in a federated data platform integrating DataSHIELD. No individual data transfer, statistics were calculated simultaneously but in parallel within each healthcare organization and only summary statistics (aggregates) were provided back to the federated data analyst.Descriptive statistics, survival analysis, regression models and correlation were first performed on the centralized approach and then reproduced on the federated approach. The results were then compared between the two approaches.ResultsThe cohort was splitted in three samples (N1 = 157 patients, N2 = 94 and N3 = 64), 11 derived variables and four types of analyses were generated. All analyses were successfully reproduced using DataSHIELD, except for one descriptive variable due to data disclosure limitation in the federated environment, showing the good capability of DataSHIELD. For descriptive statistics, exactly equivalent results were found for the federated and centralized approaches, except some differences for position measures. Estimates of univariate regression models were similar, with a loss of accuracy observed for multivariate models due to source database variability.ConclusionOur project showed a practical implementation and use case of a real-world federated approach using DataSHIELD. The capability and accuracy of common data manipulation and analysis were satisfying, and the flexibility of the tool enabled the production of a variety of analyses while preserving the privacy of individual data. The DataSHIELD forum was also a practical source of information and support. In order to find the right balance between privacy and accuracy of the analysis, set-up of privacy requirements should be established prior to the start of the analysis, as well as a data quality review of the participating healthcare organization.