Find research datasets worth reusing
Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.
2,639
datasets available to search
ShareScore release 0.9.0
Dataset results
2,639 results for “Robotic”
Software system integration of heterogeneous swarms of robots
<p>Paper: Software system integration of heterogeneous swarms of robots</p> <p>Robots> E-puck, E-puck+Raspberry Pi 2 B, Khepera III, Khepera III+Raspberry Pi 2 B, Koala+Laptop.</p> <p>Using the ROS drivers a series of tests were devised to prove that the robots in the swarm can communicate at a software level using ROS. The robots have individual goals, in this case to follow simple movement commands. The robots move after receiving command information from their program whether it is local or from another robot. All programs needed for each test will be called individually. A desktop computer together with the programs SSH and X11VNC were used in the tests to connect remotely to the robots, i.e. the laptop and Raspberry PIs mounted on the robots.</p> <p><em>1. Single-Minded Robot Swarm.</em> A main robot Koala is running all the drivers and programs for each robot. The programs for the robots communicate through the roscore and send commands to their respective robots using Bluetooth.</p> <p><em>2. Swarm with Leader Robot and Standalone Basic Robots. </em>This test is done with three Standalone robots. Each robot runs it’s own drivers and they all connect to a single roscore. In this test a Koala robot is used as a Leader running roscore and rtab map program for ROS (http://wiki.ros.org/rtabmap_ros) to create maps with the Kinect sensor. The mapping information is needed for SLAM. To connect to a single roscore all the robots must have the following two ROS commands run in the terminal to tell the system where the roscore is run and the IP of each robot: <em>export ROS_MASTER_URI=http://Leader_IP:11311 </em>and<em> export ROS_IP=Current_Robot_IP .</em></p> <p><em>3. Swarm with Standalone Robots Connecting to a Central Robot. A</em>ll the robots run their drivers and other high-level programs standalone, but they connect and communicate with each other through a single roscore running on a single robot. In this test, the E-puck robot runs along with its driver and uses gmapping ROS nodes (http://wiki.ros.org/gmapping) for SLAM. The K3 is running the driver and a simple movement program. The Koala robot runs rtab map with the Kinect and has a wireless controller connected via USB for movement commands. The controller Linux drivers are linked to a ROS joy node (http://wiki.ros.org/joy) which will publish the controller data. To prove the communication between robots, the controller on the Koala is used to send movement commands to the Koala and E-puck.</p> <p><em>4. Swarm with Standalone Robots. </em>In this test each robot is completely independent, running roscore and all programs. The robots connect with each other using rosmultimaster nodes (http://wiki.ros.org/multimaster_fkie). All the robots in this test subscribe to the joy node running on the Koala, and thus move receiving commands from the controller.</p> <p><em>5. Combined Swarm - Single-Minded Robot Swarm(s) + Standalone Robots. </em>In the final test all Standalone robots run their own roscore and the rosmultimaster node. The Standalone robots have other simple robots connected to them using Bluetooth and run their own drivers, other more advanced programs and the simple robots’ ROS drivers. A tested goal here is to have simple robots such as those from the first test receive commands from another Standalone robots in the swarm through its Leader robot. The Koala controller joy node was used again publishing data on the ROS network. Any robot driver connected to this network can subscribe to the publishing joy node and thus the simple robots not connected to the Koala with the controller, receive commands from the controller through their own respective Leader robot.</p>
Autonomous Mobile Robots: Past, Present and Future of SLAM 2013
<p>“Autonomous Mobile Robots: Past, Present and Future of SLAM” 2013. In:Workshop<br /> at the First RSI/ISM International Conference on Robotics and Mechatronics by<br /> Sharif University of Technology. Presenter: Prof. Hamid D. Taghirad, 2013.</p>
P colonies and P swarms for controlling robot swarms. Experimental setups and demonstration videos
<p>These eleven videos present different experimental scenarios used to test the flexibility and functioning of the LULU P colony/P swarm simulator and of the associated application Lulu_Kilobot for controlling robot swarms. Both applications will be published on Github under an open-source license. The input P colony (input) file, swarm configuration (config) file and V-REP scene (.ttt) are available for each video in the associated .zip archive.</p> <p>The LULU simulator was included as a Python module in Lulu_Kilobot in order to test robot controllers based on P colonies, XP colonies, and P swarms for swarms of up to 10 Kilobot robots.</p> <p>In the following sections, we present a small description for each of the eleven attached videos.</p> <p>-----------------------------------------------------------------------------------------------<br /> 1_clone_10_circle</p> <p>This video demonstrates the use of the robot cloning function of the vrep_bridge script in order to create 9 distinct copies of the source robot and distribute them on a circle around the source robot. The copies are so positioned by a distribution function that can be adapted to other forms. This cloning function allows one to generate large swarms of robots with ease.</p> <p>-----------------------------------------------------------------------------------------------<br /> 2_one_pcolony_for_three_kilobots</p> <p>In this video, we simulate a simple P minus colony using Lulu_Kilobot, on three different robots. At each subtraction, the robots move one step forward. At the beginning of the clip one can see the Robot - P colony association table, where each robot has a distinct copy of the original P colony.</p> <p>-----------------------------------------------------------------------------------------------<br /> 3_pswarm_5_robots_3_colonies</p> <p>This video demonstrates the flexibility offered by the config file of Lulu_Kilobot. From the config file we explicitly specify that the first two robots should use the go straight P colony. For the other colonies, we specify the number of robots that should be assigned, go left = 1 and go right = 2.</p> <p>From the Robot - P colony association table, one can see that the first robot that is assigned a P colony uses the original P colony while the others use an independent copy of the P colony.</p> <p>-----------------------------------------------------------------------------------------------<br /> 4_pswarm_2_robots_avoid_collision</p> <p>In this experiment, we test the msg_distance agent from the input module, by continuously checking the distance from another robot.</p> <p>If the distance is short, then we stop the movement and otherwise continue to subtract f objects from the environment and move forward.</p> <p>Each of the two robots has a different P colony that was designed to check the distance from the other robot (robot_0 checks the distance from robot_1).</p> <p>-----------------------------------------------------------------------------------------------<br /> 5_pswarm_2_robots_xp_colonies_15_steps</p> <p>In this video we employ the exteroceptive communication rules (denoted by <=>) in order to synchronize the movement of two robots. The first robot moves forward 15 steps and after it stops, it signals the second robot to start moving. At this signal, the second robot starts to turn left 15 steps.</p> <p>This shows the utility of exteroceptive rules that allow XP colonies (P colonies with exteroceptive rules) to communicate using the global P swarm environment.</p> <p>-----------------------------------------------------------------------------------------------<br /> 6_1_pswarm_10_robots_disperse_steps_infinite_loop</p> <p>In this video we run a more complex algorithm that involves the use of the following modules: msg_distance, led_rgb, and motion.</p> <p>This video demonstrates dispersion, which is a typical self-deploying scenario in swarm robotics. The robots should position themselves away from one another, so that each robot is at least at a minimum distance from each of its neighbours.</p> <p>All decisions are taken by the command module, on the basis of the received input data from msg_distance. A new direction of motion (and color) is randomly chosen if there are other robots closer than a pre-set threshold distance.</p> <p>-----------------------------------------------------------------------------------------------<br /> 6_2_pswarm_10_robots_optimized_disperse_infinite_loop</p> <p>This video is an optimized version of 6.1 (10 robots disperse).</p> <p>The optimization consists in only exchanging data with V-REP when a new input request is detected in the input agents or likewise a new command object is detected in the output agents.</p> <p>This results in a step-less movement of the robots and reduces the time needed for a new decision to be applied resulting in a faster overall simulation time.</p> <p>-----------------------------------------------------------------------------------------------<br /> 7_pswarm_10_robots_optimized_disperse_infinite_loop_with_intruder</p> <p>In this video we use the previously presented optimized dispersion algorithm (6.2) and introduce an intruder robot into the scene in order to evaluate the influence that this intruder has over the behaviour of the swarm.</p> <p>One can see that robots that have stopped their movement, restart dispersing when the intruder robot is brought close enough. This can cause a chain reaction and ultimately cause the swarm to reposition.</p> <p>-----------------------------------------------------------------------------------------------<br /> 8_1_pswarm_10_robots_secure_disperse_fast</p> <p>In this video we test the proposed security protocol (based on entity authentification and P colony based id check) using the non-optimized dispersion algorithm.</p> <p>In this film we see that the robots ignore the intruder robot even though it is placed in the middle of the swarm. On the other hand, if we bring a swarm member robot close to another swarm member robot, these two will start to disperse normally.</p> <p>-----------------------------------------------------------------------------------------------<br /> 8_2_pswarm_10_robots_secure_disperse_infinite_loop_2</p> <p>In this video we present the optimized (see 6.2 for details) version of the secured dispersion algorithm.</p> <p>As was the case of the secured un-optimized version (8.1), in this secured version we test the influence of the intruder on the behaviour of the swarm by moving the intruder close to the center of the swarm and also moving the intruder close to the swarm after the dispersion is finished. We also note that if two member robots approach, the algorithm continues to work normally.</p> <p>-----------------------------------------------------------------------------------------------<br /> 9_pswarm_10_robots_secure_d_min_disperse_infinite_loop</p> <p>In this clip we show the effects of transparent input data processing.</p> <p>The command agent always requests the smallest distance available from the neighbour list, by using the d_min command.</p> <p>When the intruder robot is the nearest robot (the smallest value in the list) d_min will always return the intruder robot which cannot be processed because it is unknown to the swarm members. For this reason, a member robot that is in this situation will be blocked by the intruder robot.</p>
3rd ACSE Robot Rescue and Search Competition
<p>X. Dai, S.A. Tafrishi, and Y. Kuang. “3rd ACSE Robot Rescue and Search Competition”. University of Sheffield, First Runner-up Team, Master Shifu Robot, May<br /> 2013.</p>
Demonstration videos on using P swarms for deployment tasks in swarm robotics.
<p>In this video we present the capability of the Lulu P colony / P swarn simulator to control a Kilobot robot that is a member of a swarm using a P colony based controller. The swarm application presented here is the dispersion from neighbouring robots, until a certain distance from all neighbours is reached.</p> <p>In the first video entitled kilombo_disperse_1000.avi, we simulated the interactions between 1000 Kilobots using the Kilombo simulator. The i5-4240 CPU used in this test allowed the simulation of 1000 robots, each one controlled by an individual P colony at a peak speed of 29 x real world speed.<br /> <br /> The source code of the P colony used for the dispersion algorithm is also included. This P colony is defined in the input file format accepted by the Lulu P colony / P swarm simulator.</p>
Robot Self-Assembly as Adaptive Growth Process
<p>Autonomous self-assembly allows to create structures and scaffolds on demand and automatically. The desired structure may be predetermined or alternatively it is the result of an artificial growth process that adapts to environmental features and to the intermediate structure itself. In a self-organizing and decentralized control approach the robots interact only locally and form the structure collectively. Designing a complete approach that allows the robot group to collectively decide on where to start the self-assembly, that adapts at runtime to environmental conditions, and that guarantees the structural stability is challenging and does not yet exist. We present an approach to self-assembly inspired by diffusion-limited aggregation that generates an adaptive structure reacting to environmental conditions in an artificial growth process. During a preparatory stage the robots collectively decide where to start the self-assembly also depending on environmental conditions.In the actual self-assembly stage, the robots create tree-like structures that grow towards light. We report the results of robot self-assembly experiments with 50 Kilobots. Our results demonstrate how an adaptive growth process can be implemented in robots. We explain how our approach will be extended to a 3-d growth process and how robot self-assembly as an open-ended adaptive growth process opens up a multiplicity of future opportunities.</p>
XGNITIVE: A Flexible Cognitive Architecture for Robots (Experiments)
<p>Dataset of experiments for XGNITIVE: A Flexible Cognitive Architecture for Robots. More information in its repository ( https://github.com/roboticslab-uc3m/xgnitive ).</p>
Postural Optimization for a Safe and Comfortable Human-Robot Interaction: Experiment Dataset
<p>In human-robot collaboration the robot's behavior impacts the worker's safety, comfort, and his acceptance of the robotic system. In this paper we address the problem of how to improve the worker's posture during human-robot collaboration. Using postural assessment techniques, and a personalized human kinematic model, we optimize the model body posture to fulfill a task while avoiding uncomfortable or unsafe postures. We then derive a robotic behavior that leads the worker towards that improved posture. We validate our approach in an experiment involving a joint task with 39 human subjects and a Baxter torso-humanoid robot.</p> <p>This repository contains the anonymized recorded data of our experiment. For all the subjects, we have included their recorded posture, using a motion capture system, and a video taken from a camera located on the robot head. Each data is divided by subjects and by tested conditions. In a separate file, we also include the result of the survey answered alongside the experiment.</p>
Todai Robot Project Pre-university Math Benchmark data
<p>The dataset used in the experiments described in Matsuzaki et al., "Can an A.I. Win a Medal in the Math<br> Olympiad? -- Benchmarking Mechanized Math on Pre-university Problems" (to appear)</p>
Hardware of the robot RiBot V2.0 (Mechanics and Electronics)
<p>The RiBot is an actuated robotic fish-lure that was designed to study the behavior of zebrafish. The design was based on the scan of a real zebrafish to produce the electronic boards and mechanics at the right scale. The robot can be controlled using Infrared RC5 signals and can beat its tail with different range of frequencies and amplitudes.</p> <p>The electronics was designed using the software Altium designer and the mechanics using CREO 3. The step file of the mechanics can be open by other types of CAD software. The firmware of the RiBot to program the electronics is described in:</p> <p>https://zenodo.org/record/830181#.WXtAIYTyipo</p> <p>The design of the lure is presented in detail in the following publication:</p> <p>F. Bonnet, N. Crot, D. Burnier and F. Mondada. Design Methods for Miniature Underwater Soft Robots. 6th IEEE RAS & EMBS International Conference on Biomedical Robotics and Biomechatronics (BioRob 2016), Singapore, Proceedings of the IEEE RAS-EMBS International Conference on Biomedical Robotics and Biomechatronics, 2016.</p> <p> </p> <p> </p>
Experimental Validation video for paper "The Critical Role of Effective Communication in Human-Robot Collaborative Assembly"
<p>Experimental Validation video for paper "The Critical Role of Effective Communication in Human-Robot Collaborative Assembly".</p><p>The video shows a collaborative manipulator executing a collaborative assembly job with the user using a natural vocal communication architecture. The experiment compares the proposed framework with the state-of-art interaction and highlight the differences.</p>
Modular Components for synthesizing Robotic Arms with CLS-CAD
<p>A set of modular components as Autodesk Fusion 360 files suitable for synthesizing robotic arms with CLS-CAD (https://github.com/tudo-seal/CLS-CAD). The files contain the type information required for synthesis as well as the necessary taxonomy. </p><p>To use these files or reproduce results, please follow the instructions available at https://github.com/tudo-seal/CLS-CAD in the "Getting Started" Section.</p>
Multimodal Agricultural Aerial and Ground Robotics Simulation Dataset
<p><strong>Dataset description</strong></p><p>This dataset was generated using an aerial robot and a ground robot in the Webots simulator with the <a href="https://github.com/opendr-eu/opendr/tree/master/projects/python/simulation">OpenDR agricultural dataset generator tool</a>.</p><p>It consists of 13980 RGB images and their semantic segmentation counterparts taken at different lighting conditions and robot positions in an agricultural field. It also includes the annotation data comprised of the class of the object, x, and y of the top left pixel of the object bounding box, and the width and height of the object bounding box. Furthermore, it includes gps and inertial unit sensor data for UAV and gps, inertial and lidar sensor data for UGV.</p><p><strong>Folder configuration</strong></p><p>The dataset contains 4 folders for different lighting conditions:</p><ul><li>noon cloudy</li><li>noon stormy</li><li>dawn cloudy</li><li>dusk</li></ul><p>Each contains UAV and UGV folders. UAV folder includes:</p><ul><li>annotations: contains segmented images in JPG files and annotations in TXT files.</li><li>camera: contains generated RGB images.</li><li>gps: contains the three-axis location of global positioning sensor saved in TXT files.</li><li>inertial unit: contains the inertial unit date in TXT files.</li></ul><p>UGV folder includes:</p><ul><li>annotations: contains segmented images in JPG files and annotations in TXT files.</li><li>front_bottom_camera: contains generated RGB images.</li><li>Hemisphere_v500: contains the three-axis location of the global positioning sensor saved in TXT files.</li><li>imu_robotti: contains the inertial unit date in TXT files.</li><li>velodyne: contains lidar data in PCD files.</li></ul><p><strong>Data format</strong></p><p>The dataset includes</p><ul><li>The inertial measurement TXT files include Euler angles in order of Roll, Pitch, and Yaw.</li><li>The GPS measurement TXT files include the robot position in x, y, and z order.</li><li>Object annotation TXT files include the class of the object, x, and y of the top left pixel of the object bounding box, and the width and height of the object bounding box at each line for the corresponding frame.</li></ul><p><strong>File naming convention</strong></p><p>Each data is named "s_i{_segmented, _annotation}.ext", where:</p><ul><li><strong>s</strong> denotes the simulated time in seconds.</li><li><strong>i</strong> denotes the index counting every 10ms of simulated time.</li><li><strong>ext</strong> denotes the extension, "jpg" for images, "pcd" for lidar, and "txt" for the rest.</li><li>Labels <strong>_segmented</strong> and <strong>_annotation </strong>appended to the name for segmentation image and object annotations, respectively.</li></ul><p>Each segmented image uses the following RGB color mapping:</p><ul><li>Tree: 0.1, 0.4, 0.0</li><li>Apple Tree: 0.85, 0.49, 0.57</li><li>Cow: 0.380, 0.220, 0.137</li><li>Sheep: 0.937, 0.921, 0.862</li><li>Fox: 0.992, 0.376, 0.086</li><li>Barn: 0.625, 0.293, 0.226</li><li>Cat: 0.870, 0.580, 0.0</li><li>Deer: 0.415, 0.364, 0.302</li><li>Human: 1.0, 0.855, 0.672</li></ul>
Machine learning driven self-discovery of the robot body morphology
<p>Conventionally, the kinematic structure of a robot is assumed to be known and data from external measuring devices are used mainly for calibration. We take an agent-centric perspective to explore whether a robot could learn its body structure by relying on scarce knowledge and depending only on unorganized proprioceptive signals. To achieve this, we analyze a mutual-information-based representation of the relationships between the proprioceptive signals, which we call proprioceptive information graphs (pi-graph), and use it to look for connections that reflect the underlying mechanical topology of the robot. We then use the inferred topology to guide the search for the morphology of the robot; i.e. the location and orientation of its joints. Results from different robots show that the correct topology and morphology can be effectively inferred from their pi-graph, regardless of the number of links and body configuration.</p>
Low voltage electrohydraulic actuators for untethered robotics
<p>Rigid robots can be precise but struggle in environments where compliance, robustness to disturbances, or energy efficiency are crucial. This has led researchers to develop biomimetic robots incorporating soft artificial muscles. Electrohydraulic actuators are promising artificial muscles that perform comparably to mammalian muscles in speed and power density. However, their operation requires several thousand volts. The high voltage leads to bulky and inefficient driving electronics. Here, we present hydraulically amplified low-voltage electrostatic (HALVE) actuators that match mammalian skeletal muscles in average power density (50.5 W kg<sup>−1</sup>) and peak strain rate (971 % s<sup>−1</sup>) at a 4.9 times lower driving voltage (1100 V) compared to the state-of-the-art. HALVE actuators are safe to touch, waterproof, and exhibit self-clearing properties. We characterize, model, and validate key performance metrics of our actuator. Finally, we demonstrate the utility of HALVE actuators on a robotic gripper and a soft robotic swimmer.</p>
Evaluating robot workspaces with performance maps
<p><strong>Evaluating robot workspaces with performance maps</strong></p> <p><em>Aline Kluge-Wilkes, Presley Demuner Reverdito</em></p> <p>The following data set was created while validating a proposed method to evaluate robot workspaces with different performance metrics. </p> <p>The underlying paper will be presented at the CIRP CATS 2024, the 10th Conference on Assembly Technology and Systems, hosted in Karlsruhe from 24-26.April.2024 and published thereafter.</p> <p>Deploying mobile robots in line-less and mobile assembly systems without predefined formations necessitates understanding the robot's capabilities and determining the tasks it can perform. The data set concerns performance maps to assess the feasibility of assembly tasks within the robot's workspace. These maps offer quantifiable metrics to compare the suitability of base placements for mobile robots concerning the feasibility of specific assembly tasks.</p> <p>Performance maps are a discretised representation of a particular robot's distribution of selected performance metrics. The current implementation focuses on calculating manipulability, dexterity, and condition number. Using a URDF robot model and task poses as input, the metrics are computed across distributed poses within the robot's workspace at a specified resolution, forming the performance map.</p> <p>The data set contains the application of the performance map to exemplary tasks and robots.</p> <p>A Design of Experiments (DoE) was conducted to investigate the influence of three predictor variables on robot manipulation performance. The predictor variables considered in the study were as follows:</p> <p>1. Resolution (R): Two levels of resolution, 0.08m and 0.12m, were examined to assess their impact on robot performance representation.</p> <p>2. Number of poses (N): Two different quantities of poses, 25 and 50, were studied to understand how they affect the representation of the robot's capabilities.</p> <p>3. Robot Model: Two distinct robot models, the ABB IRB1600-1.2 and the ABB IRB120-0.58, were chosen for the experiments.</p> <p> </p> <p>Eight experiments were designed and executed, each with a unique combination of the predictor variables. These experiments are summarised in the Excel sheet "DOE-tests-and-results". The first table overviews the eight experiments. The following eight tables indicate the positions and the results: the respective calculated robot's performance metrics. </p> <p>The .json files specify the poses for which the respective performance metrics (dexterity index, reciprocal condition number, and manipulability index) per robot were calculated.</p> <p>The included .png files visualise the distribution of each index for each conducted experiment in a performance map. The maps were created with Matplotlib. </p> <p> </p> <p>As a comparison, the "capability maps" of both robot models with both resolutions --- generated with the ROS (Robot Operating System) library Reuleaux (http://wiki.ros.org/reuleaux) --- are provided. The generated files are given in .h5 format. The respective files are named as "CM_...". </p> <p>-------------------------------------------------------------------------------------------------------------------------------------------------------</p> <p>Acknowledgement:</p> <p>Funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany's Excellence Strategy --- EXC-2023 Internet of Production --- 390621612.</p> <p><br><br> </p>
COgITOR Soft Robot
<p>This movie features two scientists at Empa in Switzerland, Physicits Dr. Loghman Jamilpanah, and Chemist Dr. Kwanele W. Kunene, who carry out materials research for the development of a soft robot. The robot project is called COgITOR, inspired by Prof. Dr. Alessandro Chiolerio, Italian Institute of Technology, Torino, and Genova.</p> <div> <div> <div> <p><strong>COgITOR</strong> is a project funded under the topic H2020-FETOPEN-2018-2020 / H2020-FETOPEN-2018-2019-2020-01 programme, aiming at developing a liquid state cybernetic system prototype. Holonomic memory and computing, pressure sensing, and energy harvesting from thermal gradients will be achieved using colloids. The prototype will be tested in extreme environments for potential space applications.</p> <p>This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 964388</p> <p>The website reflects only the author's view and that the European Commission is not responsible for any use that may be made of the information it contains</p> </div> </div> </div> <div> <div> <div> </div> </div> </div>
Wireless flow-powered miniature robot capable of traversing tubular structures
<p>Wireless millimeter-scale robots capable of navigating through fluid-flowing tubular structures hold substantial potential for inspection, maintenance, or repair use in nuclear, industrial, and medical applications. However, prevalent reliance on external power constrains their operational range and applicable environments. Alternatives with onboard powering must trade off size, functionality, and operation duration. Here, we propose a wireless millimeter-scale wheeled robot capable of using environmental flows to power and actuate its long-distance locomotion through complex pipelines. The flow-powering module can convert flow energy into mechanical energy, achieving an impeller speed of up to 9595 revolutions per minute, accompanied by an output power density of 11.7 watts per cubic meter and an efficiency of 33.7%. A miniature gearbox module can further transmit the converted mechanical energy into the robot's locomotion system, allowing the robot to move against water flow at an average rate of up to 1.05 meters per second. The robot's motion status (moving against/with flow or pausing) can be switched using an external magnetic field or an onboard mechanical regulator, contingent upon different proposed control designs. Additionally, we design kirigami-based soft wheels for adaptive locomotion. The robot can move against flows of various substances within pipes featuring complex geometries and diverse materials. Solely powered by flow, the robot can transport cylindrical payloads with a diameter of up to 55% of the pipe's diameter and carry devices, such as an endoscopic camera for pipeline inspection, a wireless temperature sensor for environmental temperature monitoring, and a leak-stopper shell for infrastructure maintenance.</p>
Causal HRSI Dataset: Human-Robot Spatial Interaction Dataset for Causal Analysis from Mobile Platforms
<h2>Causal HRSI Dataset: Human-Robot Spatial Interaction Dataset for Causal Analysis from Mobile Platforms</h2> <div>The dataset captures a Human-Robot Spatial Interaction (HRSI) scenario between a person and the TIAGo robot. It focuses specifically on human-goal and human-robot spatial interaction in an indoor environment, captured from the perspective of a 3D Velodyne VLP-16 LiDAR mounted on the TIAGo robot. It includes:</div> <ul> <li>rosbags containing: Velodyne LiDAR point clound, robot and human state (position, orientation and velocities);</li> <li>CSV files containing trajectories of the person and the robot generated by post-processing the rosbags;</li> <li>the map of the environment extracted from the TIAGo robot.</li> </ul> <p><strong>15 participants</strong> took part in the experiment, with the dataset capturing <strong>5 minutes of HRSI motion for each participant</strong>.</p> <h3>Experiment Description</h3> <p>The experiment and data collection occurred in a laboratory room of the University of Lincoln (UK), measuring 5 x 8.2m. <br>Fifteen participants (6 females, aged between 25 and 55) took part in the experiment. Seven of them were used to work with a robot. They were required to walk between four goal positions and avoid the robot if a cross occurs. A predefined rectangular path was set for the TIAGo robot to navigate along the room and generate frequent interactions with the participants.</p> <p>The experimental procedure can be described as follows. Each participant started from one of the four target positions. The next target position was randomly chosen by the participant, who then started moving towards it. Upon reaching the goal position, the participant stopped there and randomly chose the next goal, repeating the process for 5 minutes. In this experimental setting, the robot was considered by the participant as an obstacle to avoid while walking towards their target positions.</p> <h3>Directory Structure</h3> <p>Dataset<br>|<br>|____Map: folder containing the map of the environment extracted from the TIAGo robot<br>|<br>|____RosBags: forder containing the rosbag for each partipant<br>| |____A1.bag<br>| |____A2.bag<br>| |____A3.bag<br>| |____A4.bag<br>| |____A5.bag<br>| |____A6.bag<br>| |____A7.bag<br>| |____A8.bag<br>| |____A9.bag<br>| |____A10.bag<br>| |____A11.bag<br>| |____A12.bag<br>| |____A13.bag<br>| |____A14.bag<br>| |____A15.bag<br>|<br>|____Trajectories: postprocessed trajectories extracted for the rosbag files <br> |____A1_traj.csv<br> |____A2_traj.csv<br> |____A3_traj.csv<br> |____A4_traj.csv<br> |____A5_traj.csv<br> |____A6_traj.csv<br> |____A7_traj.csv<br> |____A8_traj.csv<br> |____A9_traj.csv<br> |____A10_traj.csv<br> |____A11_traj.csv<br> |____A12_traj.csv<br> |____A13_traj.csv<br> |____A14_traj.csv<br> |____A15_traj.csv</p>
Robot Arm
<p>This is a pre-processed version of the dataset saved in numpy format.</p> <p>The data are 3-dimensional arrays of shape [n_samples, time_steps, n_variables]. The data can be loaded as follows:</p> <pre><code>loaded_data = np.load("Robot.npz") Xtr = loaded_data['Xtr'] # Training data of shape (100, 15, 6) Ytr = loaded_data['Ytr'] # Training labels of shape (100, 1) Xte = loaded_data['Xte'] # Test data of shape (64, 15, 6) Yte = loaded_data['Yte'] # Test labels of shape (64, 1)</code></pre>
ScienceDex guides
Understand access before you commit
These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.
Allen Brain Atlas
Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.
Annotated Behaviour and Observability Dataset (ABODe)
ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.
DANDI Archive for NWB datasets
DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.
International Brain Laboratory public data
The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.
OpenNeuro
OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.