Robot Packing With Known Items and Nondeterministic Arrival Order

Abstract

This article formulates two variants of packing problems in which the set of items is known, but the arrival order is unknown. The goal is to certify that the items can be packed in a given container and/or to optimize the size or cost of a container so that that the items are guaranteed to be packable, regardless of arrival order. The nondeterministically ordered packing (NDOP) variant asks to generate a certificate that a packing plan exists for every ordering of items. Quasionline packing (QOP) asks to generate a partially observable packing policy that chooses the location of each item as their arrival order is revealed one-by-one, with all of the remaining items certified to be packable regardless of their future arrival order. Theoretical analysis demonstrates that even the simple subproblem of verifying the feasibility of a packing policy is NPcomplete. Despite this worst case complexity, practical solvers for both NDOP and QOP are developed. Multiple extensions to the basic nondeterministic problem are presented, including packing with a fixed-capacity buffer and packing with equivalent objects. Experiments demonstrate that these algorithms can be applied to packing irregular 3-D shapes with stability and manipulator loading constraints. Note to Practitioners-Automatic packing of objects into containers, such as a shipping box, has applications in warehouse automation for e-commerce applications and less-than-truckload shipping. In many scenarios, a container needs to be preselected for a set of objects before they arrive, and the order of the objects cannot be controlled. This article studies the theoretical foundations of packing problems, in which the set of items is known, but the arrival order is unknown. The algorithms presented in this article can certify whether a container can hold a set of objects in the case where the arrival order is revealed at once, one-by-one, and where a limited set of objects can be put aside in a buffer before packing. These algorithms can be used to choose optimal containers and packing strategies, both for robot and human packers.

Robotic Grasp/Motion Planning for Efficient Packing Assuming Multiple Grippers

“…Y. Wu et al [12] used the genetic algorithm to solve a threedimensional bin-packing problem, considering the orientation of each object to determine the position of each container. Wang and K. Hauser [13] proposed a method for inserting irregular three-dimensional objects into containers, independent of the arrival order. A. Yasuda et al [14] applied the particle swarm optimization method to determine the position of each object offline and used a robot to perform the packing , while V. Chekanin [15] and R. Sridhar [16] showed that the genetic algorithm, which guarantees the speed and packing rate to a certain extent, is optimal for the packing problem.…”

Fast High-Quality Tabletop Rearrangement in Bounded Workspace

“…More recently, some of the authors [2] examined an optimization objective, running buffers, which is the size of the external space needed for the rearrangement task, and also examined an unlabeled setting. Similar graph structures are also used in other robotics problems, such as packing problems [19]. Deep neural networks have been also applied to detect the embedded dependency graph of objects in a cluttered environment to determine the ordering of object retrieval [9].…”

Task-driven Perception and Manipulation for Constrained Placement of Unknown Objects

“…This section discusses existing pick-and-place manipulation pipelines, object representations for these pipelines and assumptions about the object's shape and category. Manipulation pipelines for pick-and-place: Given access to object models, previous work has addressed problems such as bin-picking [9], tight-packing [1], [2] and placement of grasped objects in clutter [3]. Most manipulation pipelines for novel objects [4], [5] focus on picking the object but do not address the problem of constrained placement.…”

“…Such scenarios occur in logistics applications, such as packing items into boxes, or in service robotics, such as inserting a book into a gap in a bookshelf. Recent work has focused on variants of this problem, such as bin-packing [1], [2] and table-top placement in clutter [3]. Nevertheless, in many cases a geometric and textured 3D model for the manipulated object is assumed to be known.…”

“…240 real-world manipulation experiments are performed to compare the proposed solution and the straightforward picksense-and-place alternative 1 . The experiments demonstrate that the proposed pipeline is both robust and efficient in handling objects with no prior models within the limitations of the end-effector and the sensor.…”