About this Research Topic
Grasping is a complex problem since it is a multi-disciplinary task that spans from the mechatronic design of grippers to higher-level domains like planning, perception, and control. Therefore, this Research Topic "Robotic Grasping and Manipulation in the Real World: Methods, Vision, Applications" aims to present the recent progress in each of the aspects that concern grasping or its applications in tackling the most relevant problems that are not solved yet. In the last few years, research in the field of robotic grasping and manipulation focused on the following areas: the design of new hybrid grippers or hands that can provide multiple grasping modalities to help handle a wide variety of objects; the implementation of advanced grasp planners, which, exploiting machine learning approaches, can detect the pose of the objects more accurately and solve some cluttered scenes; the integration of advanced sensors inside robotics grippers to achieve fine control of the grasping phase or improve in-hand manipulation capabilities; the use of smart materials to realize soft grippers for delicately grasping fragile objects; the design of new grasp quality metrics or state models for soft or deformable objects; the design of benchmarking protocols to allow a systematic comparison of the performance of different robotic solutions.
Topics of interest for this Research Topic include but are not limited to:
- modeling, design, and control methods for robotic grippers and hands
- grasping pipelines for known and unknown objects
- perception methods for grasping and dexterous manipulation in cluttered environments
- deep learning methods for robust grasping
- real or synthetic photorealistic dataset for machine vision and robot perception for grasping synthesis
- benchmarks and evaluation procedures for performance comparison
- approaches for handling fragile and deformable objects
- quality metrics that can be applied to fragile and deformable objects
- approaches for advanced manipulation tasks such as in-hand manipulation or slippage detection
- robot programming through imitation learning for manipulation tasks
Review papers that address in detail the existing state-of-the-art solutions proposed in recent years for one of the aforementioned points are also welcome.
Salvatore D'Avella is with the Mechanical Intelligence Institute, Department of Excellence in Robotics and AI, Sant'Anna School of Advanced Studies, Pisa, Italy. His current research interest is Robotics and AI for industry 4.0 and logistics, mainly focused on autonomous grasping and manipulation.
Keywords: autonomous grasping, dexterous manipulation, multi-modal grasping, robotic perception, sim-to-real transfer, deep learning, reinforcement learning, self-supervised learning, deformable objects, tactile sensing, sensors fusion, soft grippers, quality metrics
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.