Learning-Based Control for Autonomous Robotic Manipulation

Dear Colleagues,

Learning-based control has emerged as a transformative paradigm in autonomous robotic manipulation, enabling robots to acquire complex manipulation skills from data, adapt to unstructured environments, and improve performance through interaction. However, significant challenges remain at the intersection of learning, control, and physical interaction. These include guaranteeing stability and safety during real-time execution, achieving generalization across diverse tasks and objects, bridging the gap between simulation and the real world, and incorporating structured knowledge and cognitive reasoning into control policies. Addressing these challenges is essential to deploy autonomous manipulators in critical domains such as manufacturing, logistics, healthcare, and space exploration.

The purpose of this Special Issue is to gather original research articles that reflect innovative methods in the following two primary tracks within the intersection of learning-based control and autonomous robotic manipulation:

Track 1: Learning for Manipulation Control focuses on advancing learning algorithms to improve the performance, adaptability, and robustness of robotic manipulation systems. Relevant areas include deep reinforcement learning for dexterous grasping, imitation learning from human demonstrations, adaptive and robust control using learned dynamics models, and real-time learning for dynamic or contact-rich tasks. This track encourages research that integrates learning with classical control theory to achieve efficient, stable, and generalizable manipulation behaviors.

Track 2: Cognitive and Knowledge-Driven Manipulation explores the integration of higher-level reasoning, knowledge representation, and structured learning into manipulation control systems. This track covers research on cognitive architectures for task and motion planning, knowledge graph-driven reasoning for manipulation under uncertainty, combining symbolic knowledge with low-level learning-based control, semantic scene understanding for object manipulation, and commonsense reasoning for long-horizon tasks. It also invites studies on knowledge transfer, lifelong learning, and the use of large language models or foundation models for robotic manipulation.

This Special Issue seeks contributions that capture theoretical, methodological, and practical advancements in both learning-enabled manipulation control and cognitive-driven robotic manipulation. While key topics are suggested for each track, the scope remains open to a broad range of scholarly perspectives that advance this interdisciplinary dialogue.

Dr. Kui Qian
Dr. Shi Wang
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 250 words) can be sent to the Editorial Office for assessment.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-anonymized peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Electronics is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• learning-based robot control
• autonomous robotic manipulation
• deep reinforcement learning for manipulation
• imitation learning from demonstration
• dexterous grasping and hand manipulation
• adaptive control for robotic arms
• robust manipulation under uncertainty
• sim-to-real transfer for robotics
• real-time learning for contact-rich tasks
• vision-language models (VLMs) for robotics
• vision-language-action (VLA) models for manipulation
• world models for robotic control
• latent dynamics learning and model-based RL
• Cognitive robotics and task planning
• knowledge-driven manipulation reasoning
• semantic scene understanding for grasping
• long-horizon manipulation planning
• safe and stable learning control
• vision-based robotic manipulation
• force and tactile feedback learning
• human–robot collaborative manipulation
• generalization in robotic manipulation