Article
Computer Science, Artificial Intelligence
Xin Shu, Fenglei Ni, Xinyang Fan, Shuai Yang, Changyuan Liu, Baoxu Tu, Yiwei Liu, Hong Liu
Summary: Humanoid robots have attracted attention for their compatibility with human environments, but they still lack stability and reliability in real-world settings. Combining humanoid robots with various stable mobile platforms is a favored solution. This article presents a versatile humanoid robot platform, which allows flexible deployment in diverse scenarios by incorporating multimodal perception and extensible interfaces. The platform has achieved impressive integration, lightness, dexterity, and strength, with the goal of human-intelligent manipulation skills for human-engineered environments. The article elaborates on design choices, subsystems, and demonstrates the platform's performance through experiments.
CAAI TRANSACTIONS ON INTELLIGENCE TECHNOLOGY
(2023)
Article
Education & Educational Research
Christina Chalmers, Therese Keane, Marie Boden, Monica Williams
Summary: This three-year study explores teachers' perceptions of humanoid robots and finds that their deployment in education can enhance student engagement. The study advocates for a constructivist approach to using humanoid robots in the curriculum.
EDUCATION AND INFORMATION TECHNOLOGIES
(2022)
Article
Computer Science, Information Systems
Ester Martinez-Martin, Angel P. Del Pobil
Summary: Robotics research has shifted from laboratories to factories and real-world environments in recent decades. This paper presents the evolution of robot vision for manipulation through analyzing various robot developments, highlighting the importance of using different techniques depending on the task and scenario at hand.
Article
Computer Science, Artificial Intelligence
S. Hamidreza Kasaei, Jorik Melsen, Floris van Beers, Christiaan Steenkist, Klemen Voncina
Summary: The development of service robots involves multiple research fields, with current technologies improving the coupling between object perception and manipulation to adapt to changing environments. While robots can now recognize objects and plan collision-free trajectories for grasping, they still rely heavily on large amounts of training data, leading to complex re-programming when environmental changes occur.
JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS
(2021)
Article
Chemistry, Analytical
Hao Guo, Meichao Song, Zhen Ding, Chunzhi Yi, Feng Jiang
Summary: Learning from visual observation is a significant challenge in robotic manipulation. This method improves policy general performance among multitasks by using a lite Transformer based visual encoder and attention mechanism. Experimental results show superior performance compared to alternative encoders and fewer training steps.
Article
Computer Science, Artificial Intelligence
Jiaji Yang, Esyin Chew, Pengcheng Liu
Summary: The study proposes a new interactive framework for humanoid service robots, integrating Trevarthen's companionship theory and neural image captioning generation algorithm, to help the robots better respond to complex social environments.
PEERJ COMPUTER SCIENCE
(2021)
Article
Computer Science, Information Systems
Matthias Kerzel, Philipp Allgeuer, Erik Strahl, Nicolas Frick, Jan-Gerrit Habekost, Manfred Eppe, Stefan Wermter
Summary: Researchers propose a novel semi-humanoid robot NICOL that can efficiently collaborate with humans in social interaction and physical tasks. They introduce the design concept of NICOL and evaluate its manipulation abilities using a new end-to-end hybrid neuro-genetic visuomotor learning approach, which outperforms existing inverse kinematics solvers.
Article
Chemistry, Multidisciplinary
Jaehoon Sim, Seungyeon Kim, Suhan Park, Sanghyun Kim, Mingon Kim, Jaeheung Park
Summary: The JET humanoid robot, developed based on the THORMANG platform, aims to be applied in industrial and service fields by increasing height, reducing weight, and expanding workspace. The simplification of maintenance process and experiments conducted validate the practicality of the design concepts.
APPLIED SCIENCES-BASEL
(2021)
Article
Automation & Control Systems
Mario Selvaggio, Akash Garg, Fabio Ruggiero, Giuseppe Oriolo, Bruno Siciliano
Summary: This article proposes a model predictive non-sliding manipulation (MPNSM) control approach for safe object transportation on a robotic manipulator. The approach enforces sticking contacts and ensures that joint positions, velocities, and input torques do not exceed physical limits. Simulation results using a torque-controlled 7-DoF KUKA LBR IIWA robotic manipulator and real experiments on a 21-DoF humanoid robotic platform validate the proposed approach.
IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY
(2023)
Article
Engineering, Electrical & Electronic
Sheeraz Athar, Gaurav Patel, Zhengtong Xu, Qiang Qiu, Yu She
Summary: This study presents an innovative robotic finger called VisTac, which combines high-resolution tactile and visual perception in a single device. Using multimodal sensing capabilities, VisTac can accurately locate objects in 3-D space and manipulate them according to tasks.
IEEE SENSORS JOURNAL
(2023)
Article
Robotics
Takumi Hara, Takashi Sato, Tetsuya Ogata, Hiromitsu Awano
Summary: In this study, we propose a haptic shared control system that predicts human manipulation intentions using a neural network and adaptively adjusts the strength of haptic guidance presentation based on the uncertainty of the inference results. The experiments conducted with the Nextage OPEN robot demonstrate a reduction of 17.1% in task execution time compared to the existing method that ignores the confidence of the neural network.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2023)
Article
Chemistry, Multidisciplinary
Matteo Russo, Marco Ceccarelli, Daniele Cafolla
Summary: This paper presents a model-based motion planning approach for cable-driven compliant mechanisms with a flexible backbone. A novel kinematic model is introduced to describe and analyze the torso mechanism for humanoid robots, improving the accuracy and efficiency of the system.
APPLIED SCIENCES-BASEL
(2021)
Article
Computer Science, Artificial Intelligence
Alexandre F. V. Muzio, Marcos R. O. A. Maximo, Takashi Yoneyama
Summary: RoboCup 3D Soccer Simulation is a robot soccer competition that uses a high-fidelity simulator and autonomous humanoid agents. This article focuses on learning humanoid robot behaviors and evaluates the effectiveness of the learned dribble policy in the Soccer 3D environment.
JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS
(2022)
Article
Computer Science, Software Engineering
Gauri Tulsulkar, Nidhi Mishra, Nadia Magnenat Thalmann, Hwee Er Lim, Mei Ping Lee, Siok Khoong Cheng
Summary: Social Assistive Robotics are increasingly used in care settings to provide psychosocial support for the elderly with cognitive impairments. However, there are shortcomings in satisfying the need for companionship. Research shows that interactions with social robots can improve the well-being of the elderly by compensating for emotional, cognitive, and psychosocial deficiencies.
Article
Chemistry, Analytical
Giovanni Mezzina, Daniela De Venuto
Summary: In this paper, a framework called MONOCULAR is proposed to add object manipulation capabilities to social robots. The framework uses RGB cameras and a 3D depth sensor to localize objects, define shelf characteristics, and adapt manipulation routines. MONOCULAR achieved high grabbing success rates for both centrally and laterally placed objects on shelves. A case study using Pepper as a drug delivery operator demonstrated the applicability of the MONOCULAR framework. The framework is designed to be fully automatic and protect sensitive user data.