Simulated and Real Robotic Reach, Grasp, and Pick-and-Place Using Combined Reinforcement Learning and Traditional Controls

Lobbezoo, Andrew; Kwon, Hyock-Ju

doi:10.3390/robotics12010012

Cited by 16 publications

(7 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The latter typically relies on implicit prediction of the same mechanics to generate accurate value estimates. These findings emphasize the superiority of explicitly integrating physical mechanics into the policy formulation process [7]. In order to address the issue regarding initial categorization with many groups, the researchers of [8] suggested a DRL method based on the (DQN) technique.…”

Section: Literature Reviewmentioning

confidence: 92%

Dynamic Object Recognition and Robot Grasping using CNN-BiLSTM with integration of Double-DQN

Soumyashree,

Singh,

Gill

2024

Preprint

View full text Add to dashboard Cite

This study introduces a novel object recognition algorithm employing the CNN-BiLSTM technique. Through the integration of advanced sensory technologies and deep reinforcement learning approaches, the proposed system dynamically adjusts its detection parameters to accommodate diverse object geometries, sizes, and surface properties. The approach entails training a neural network agent to discern, from an image window, which objects within a predetermined region warrant focused attention. To effectively handle high-dimensional and time-series data, a Double DQN framework is utilized. The proposed approach amalgamates object recognition algorithms with sophisticated reinforcement learning techniques, notably Double DQN, enabling robots to dynamically perceive and grasp items in varied environments. The outcomes were obtained by conducting comparative research to evaluate the effectiveness of current approaches in comparison to the proposed method. The Q-network is trained to execute two distinct motions, namely "twist" and "push," with the intended motion provided as an assigned parameter (+1 for "push" and-1 for "twist"). Moreover, the network can generate intermediate movements between the learned motions when the task parameter falls between-1 and +1. The findings reveal notable enhancements in grasping success rates and adaptability across diverse item sizes, shapes, and conditions. Notably, the study reports a significant increase in the task completion rate for grasping, achieving a success rate of 90%.

show abstract

Section: Literature Reviewmentioning

confidence: 92%

Dynamic Object Recognition and Robot Grasping using CNN-BiLSTM with integration of Double-DQN

Soumyashree,

Singh,

Gill

2024

Preprint

View full text Add to dashboard Cite

show abstract

“…In [ 77 ], the authors introduce a novel pipeline that combines traditional control and reinforcement learning (RL) techniques for both simulated and real-world environments to validate RL methods across various scenarios, including reach, grasp, and pick-and-place tasks. Two algorithms, Soft Actor-Critic (SAC) and Proximal Policy Optimization (PPO), are employed in this study.…”

Section: Rl Applicationmentioning

confidence: 99%

“…For safety concerns, which could involve objects or the robot itself, avoiding damages is crucial, limiting the number of samples that can be collected. Moreover, the disparity between the simulation and the real-world environment remains a challenge, and most recent studies have faced this issue [ 75 , 76 , 77 ]. The samples used in the simulation environment could lead to a good policy that may not transfer well to the real world due to variations in the samples, necessitating the collection of more samples for fine-tuning.…”

Section: Challenges Conclusion and Future Directionsmentioning

confidence: 99%

Reinforcement Learning Algorithms and Applications in Healthcare and Robotics: A Comprehensive and Systematic Review

Al-Hamadani,

Fadhel,

Alzubaidi

et al. 2024

Sensors

View full text Add to dashboard Cite

Reinforcement learning (RL) has emerged as a dynamic and transformative paradigm in artificial intelligence, offering the promise of intelligent decision-making in complex and dynamic environments. This unique feature enables RL to address sequential decision-making problems with simultaneous sampling, evaluation, and feedback. As a result, RL techniques have become suitable candidates for developing powerful solutions in various domains. In this study, we present a comprehensive and systematic review of RL algorithms and applications. This review commences with an exploration of the foundations of RL and proceeds to examine each algorithm in detail, concluding with a comparative analysis of RL algorithms based on several criteria. This review then extends to two key applications of RL: robotics and healthcare. In robotics manipulation, RL enhances precision and adaptability in tasks such as object grasping and autonomous learning. In healthcare, this review turns its focus to the realm of cell growth problems, clarifying how RL has provided a data-driven approach for optimizing the growth of cell cultures and the development of therapeutic solutions. This review offers a comprehensive overview, shedding light on the evolving landscape of RL and its potential in two diverse yet interconnected fields.

show abstract

“…Combining artificial intelligence and experience sharing in LFD is also one of the research hotspots in this field in recent years [32][33][34][35][36][37][38][39]. Kehoe et al [40] proposed an algorithm based on deep neural networks.…”

Section: Related Workmentioning

confidence: 99%

Research on LFD System of Humanoid Dual-Arm Robot

Cui,

Kou,

Chen

et al. 2024

Symmetry

View full text Add to dashboard Cite

Although robots have been widely used in a variety of fields, the idea of enabling them to perform multiple tasks in the same way that humans do remains a difficulty. To solve this, we investigate the learning from demonstration (LFD) system with our independently designed symmetrical humanoid dual-arm robot. We present a novel action feature matching algorithm. This algorithm accurately transforms human demonstration data into task models that robots can directly execute, considerably improving LFD’s generalization capabilities. In our studies, we used motion capture cameras to capture human demonstration actions, which included combinations of simple actions (the action layer) and a succession of complicated operational tasks (the task layer). For the action layer data, we employed Gaussian mixture models (GMM) for processing and constructing an action primitive library. As for the task layer data, we created a “keyframe” segmentation method to transform this data into a series of action primitives and build another action primitive library. Guided by our algorithm, the robot successfully imitated complex human tasks. Results show its excellent task learning and execution, providing an effective solution for robots to learn from human demonstrations and significantly advancing robot technology.

show abstract

Simulated and Real Robotic Reach, Grasp, and Pick-and-Place Using Combined Reinforcement Learning and Traditional Controls

Cited by 16 publications

References 22 publications

Dynamic Object Recognition and Robot Grasping using CNN-BiLSTM with integration of Double-DQN

Dynamic Object Recognition and Robot Grasping using CNN-BiLSTM with integration of Double-DQN

Reinforcement Learning Algorithms and Applications in Healthcare and Robotics: A Comprehensive and Systematic Review

Research on LFD System of Humanoid Dual-Arm Robot

Contact Info

Product

Resources

About