Effective motion planning is an indispensable prerequisite for the optimal performance of robotic manipulators in any task. In this regard, the research and application of reinforcement learning in robotic manipulators for motion planning have gained great relevance in recent years. The ability of reinforcement learning agents to adapt to variable environments, especially those featuring dynamic obstacles, has propelled their increasing application in this domain. Notwithstanding, a clear need remains for a resource that critically examines the progress, challenges, and future directions of this machine learning control technique in motion planning. This article undertakes a comprehensive review of the landscape of reinforcement learning, offering a retrospective analysis of its application in motion planning from 2018 to the present. The exploration extends to the trends associated with reinforcement learning in the context of serial manipulators and motion planning, as well as the various technological challenges currently presented by this machine learning control technique. The overarching objective of this review is to serve as a valuable resource for the robotics community, facilitating the ongoing development of systems controlled by reinforcement learning. By delving into the primary challenges intrinsic to this technology, the review seeks to enhance the understanding of reinforcement learning’s role in motion planning and provides insights that may suggest future research directions in this domain.

The introduction of collaborative robots in industrial environments reinforces the need to provide these robots with better cognition to accomplish their tasks while fostering worker safety without entering into safety shutdowns that reduce workflow and production times. This paper presents a novel strategy that combines the execution of contact-rich tasks, namely disassembly, with real-time collision avoidance through machine learning for safe human-robot interaction. Specifically, a goal-conditioned reinforcement learning approach is proposed, in which the removal direction of a peg, of varying friction, tolerance, and orientation, is subject to the location of a human collaborator with respect to a 7-degree-of-freedom manipulator at each time step. For this purpose, the suitability of three state-of-the-art actor-critic algorithms is evaluated, and results from simulation and real-world experiments are presented. In reality, the policy’s deployment is achieved through a new scalable multi-control framework that allows a direct transfer of the control policy to the robot and reduces response times. The results show the effectiveness, generalization, and transferability of the proposed approach with two collaborative robots against static and dynamic obstacles, leveraging the set of available solutions in non-monotonic tasks to avoid a potential collision with the human worker.

This research proposes an optimal robotic arm speed shape in neurological surgery to minimise a cost functional that uses an adaptive scheme to determine the brain tissue force. Until now, there have been no studies or theories on the shape of the robotic arm speed in such a context. The authors have applied a robotic arm with optimal speed control in neurological surgery. The results of this research are as follows: In this article, the authors propose a control scheme that minimises a cost functional which depends on the position error, trajectory speed and brain tissue force. This work allowed us to achieve an optimal speed shape or trajectory to reduce brain retraction damage during surgery. The authors have reached two main conclusions. The first is that optimal control techniques are very well suited for robotic control of neurological surgery. The second conclusion is that several studies on functional cost parameters are needed to achieve the best trajectory speed of the robotic arm. These studies could attempt to optimise the functional cost parameters and provide a mechanical characterisation of brain tissue based on real data.

In recent years, the application of AI has expanded rapidly across various fields. However, it has faced challenges in establishing a foothold in medicine, particularly in invasive medical procedures. Medical algorithms and devices must meet strict regulatory standards before they can be approved for use on humans. Additionally, medical robots are often custom-built, leading to high costs. This paper introduces a cost-effective brain retraction robot designed to perform brain retraction procedures. The robot is trained, specifically the Deep Deterministic Policy Gradient (DDPG) algorithm, using reinforcement learning techniques with a brain contact model, offering a more affordable solution for such delicate tasks.

Accurate energy prediction is essential for energy-aware planning and navigation of autonomous mobile robots (AMRs). This study investigates whether a compact feed-forward neural network (FFNN) can predict relative energy consumption from operational variables with high accuracy and interpretability. Using a curated dataset of controlled translation and rotation trials on a KUKA KMP 1500P, energy demand is expressed as the per-trial reduction in battery state of charge (SoC), defined as $N_{1\mathrm{\%}}$. For unit-free reporting, it is also considered the normalized SoC consumed per trial, defined as $E=1/N_{1\mathrm{\%}}$. Model development followed a two-stage optimization pipeline, (i) systematic feature-subset screening and (ii) cross-validated architecture and regularization search with early stopping, assessed by a composite of MSE, MAE, ${\mathrm{R}}^2$, and the 68th-percentile absolute error ($∆X_{68}$) as the prediction precision index. The selected FFNN (ReLU multilayer perceptron with L2 weight decay) achieved strong generalization on the independent test set (MAE = 0.9954, MSE = 4.5512, ${\mathrm{R}}^2$ = 0.9795, $∆X_{68}$ = 0.0193). Post hoc explainability methods (SHAP and input perturbation) identified angular velocity and linear acceleration as the dominant predictors, with payload mass exerting secondary effects. These results demonstrate that a compact, regularized FFNN provides accurate, repeatable, and interpretable energy predictions suitable for integration into digital twin platforms and downstream industrial scheduling.