Towards Robust Product Packing with a Minimalistic End-Effector

Abstract

Advances in sensor technologies, object detection algorithms, planning frameworks and hardware designs have motivated the deployment of robots in warehouse automation. A variety of such applications, like order fulfillment or packing tasks, require picking objects from unstructured piles and carefully arranging them in bins or containers. Desirable solutions need to be low-cost, easily deployable and controllable, making minimalistic hardware choices desirable. The challenge in designing an effective solution to this problem relates to appropriately integrating multiple components, so as to achieve a robust pipeline that minimizes failure conditions. The current work proposes a complete pipeline for solving such packing tasks, given access only to RGB-D data and a single robot arm with a vacuum-based end-effector, which is also used as a pushing finger. To achieve the desired level of robustness, three key manipulation primitives are identified, which take advantage of the environment and simple operations to successfully pack multiple cubic objects. The overall approach is demonstrated to be robust to execution and perception errors. The impact of each manipulation primitive is evaluated by considering different versions of the proposed pipeline, which incrementally introduce reasoning about object poses and corrective manipulation actions.

Robot Packing With Known Items and Nondeterministic Arrival Order

“…Due to recent advances in systems for warehouse automation [1], [12], researchers have been turning to new considerations that arise when implementing packing in robotic workcells. Shome et al [13] present a system for packing identical rectilinear objects that uses vision feedback and corrective manipulations to address errors that occur during tight packing. den Boef et al [14] formulate constraints to avoid collisions with a robot with a vacuum gripper during packing.…”

Learning image-based Receding Horizon Planning for manipulation in clutter

“…Working with RGB-D images, they use several neural network models (one per action primitive) to evaluate pixel-wise affordances for the corresponding actions, then execute the action at the location and orientation of the highest affordance. Also using RGB-D images, Shome et al [52] use a suction cap end-effector to perform one of three manipulation primitives, namely toppling, pulling, and pushing for bin packing tasks of cuboidal objects. For grasping in a tightly packed cluster of objects, Zeng et al [61] learns synergetic push and grasp actions over pixel-wise action-value heat map to disperse the clutter then grasp one of the objects in the scene.…”

Planning for Dexterous Ungrasping: Secure Ungrasping through Dexterous Manipulation

“…Our challenge tasks in this work are different from the traditional peg insertion in that dexterity is of critical importance. Robotic bin packing [19] is another relevant problem that addresses packing objects into a possibly confined space in an orderly manner. [20] (and references therein) presents a learning-based approach to context-aware object placement.…”