Deep Q-Learning in Robotics: Improvement of Accuracy and Repeatability

Abstract

Recent industrial robotics covers a broad part of the manufacturing spectrum and other human everyday life applications; the performance of these devices has become increasingly important. Positioning accuracy and repeatability, as well as operating speed, are essential in any industrial robotics application. Robot positioning errors are complex due to the extensive combination of their sources and cannot be compensated for using conventional methods. Some robot positioning errors can be compensated for only using machine learning (ML) procedures. Reinforced machine learning increases the robot’s positioning accuracy and expands its implementation capabilities. The provided methodology presents an easy and focused approach for industrial in situ robot position adjustment in real-time during production setup or readjustment cases. The scientific value of this approach is a methodology using an ML procedure without huge external datasets for the procedure and extensive computing facilities. This paper presents a deep q-learning algorithm applied to improve the positioning accuracy of an articulated KUKA youBot robot during operation. A significant improvement of the positioning accuracy was achieved approximately after 260 iterations in the online mode and initial simulation of the ML procedure.

Enhancing the Efficiency of a Cybersecurity Operations Center Using Biomimetic Algorithms Empowered by Deep Q-Learning

“…In autonomous vehicles [ 49 ], DQL developed systems that process vast sensory data to make real-time navigational decisions, significantly enhancing road safety and efficiency. In the realm of robotics, an application of a Deep Q-Learning algorithm to enhance the positioning accuracy of an industrial robot was studied in [ 50 ]. Additionally, DQL’s strategic decision-making prowess was applied in healthcare to optimize the security and privacy of healthcare data in IoT systems, focusing on authentication, malware, and DDoS attack mitigation, and evaluating performance through metrics like energy consumption and accuracy [ 51 ].…”

Robotics and Aseptic Processing in View of Regulatory Requirements

“…In particular, the resolution is defined as the measurement of the smallest deviation of the magnitude being measured by the sensors and therefore depends on the sensitivity of the sensors of the robot. The accuracy is defined as the maximum position error the robot performs in reaching a given assigned target point in space, while the repeatability defines and measures the robot’s ability to return to the same point in Cartesian space by always replicating the same motions [ 31 , 32 ]. The latter is the most significant parameter because traditional robotic applications are essentially designed to reach points in space in a repetitive way.…”

“…the maximum position error the robot performs in reaching a given assigned target poin in space, while the repeatability defines and measures the robot s ability to return to the same point in Cartesian space by always replicating the same motions [31,32]. The latte is the most significant parameter because traditional robotic applications are essentially designed to reach points in space in a repetitive way.…”

Modeling and Control of a Spherical Robot in the CoppeliaSim Simulator

“…The system learns by acting on the environment while operating in real-time. This is an advantage of traditional optimization methods that rely on a mathematical model and are tuned backwards in time [31,32].…”