Assembly Coplanner: co-operative assembly planner based on subassembly extraction

Abstract

The distribution of assembiy workstations enables assembly operations to be done in parallel, while the multiple routing of parts in flexible assembly systems allows the opportunistic scheduling of assembly operations. This paper presents an assembly planning system, called the Assembly Coplanner, which automatically constructs an assembly partial order and generates a set of assembly instructions from a liaison graph representation of an assembly based on the extraction of preferred subassemblies. Assembly planning in Coplanner is carried out by the co-operation of multiple planning agents, such as the geometric reasoner, the physical reasoner, the resource manager and the plan coordinator, under the constraints of finding a cost-effective assembly plan in a flexible assembly system. The Coplanner identifies spatial parallelism in assembly as a means of constructing temporal parallelism among assembly operations. This is achieved in the following way: (1) the selection of a set of tentative subassemblies by decomposing a liaison graph into a set of subgraphs based on feasibility and difficulty of disassembly; (2) the evaluation of each of the tentative subassemblies in terms of assembly cost represented by subassembly selection indices; and (3) the construction of a hierarchical partial order graph (HPOG) as an assembly plan. A case study applying the Coplanner to a mechanical assembly is illustrated in this paper.

Review and taxonomies of assembly and disassembly path planning problems and approaches

“…Blocking Graph [18] Removing parts of blocks assembly (2,4,5) puz C S M L C [19] Removing toleranced parts from an engine (6) ind C NS M L I [33] Removing parts of a model aircraft combustion engine (42) ind C S M L C [36] Removing parts of a model aircraft engine (13) ind C S M L C [68] Removing parts of a crate (5) ind C S M L C [69] N/A N/A C S M L C [79] Removing parts of cubes assembly (4) puz C S M L C Removing parts of: electric bell (22), model aircraft combustion engine (42), transmission (21), friction testing (skin) machine (36) a crate (5) ind [81] Removing parts of a latch (3) puz C S M L C Removing parts of: an electric bell (22), friction testing (skin) machine (36), a model of aircraft combustion engine (42) ind AND/OR Graph [84] Removing parts of a simple assembly (4) puz C S M NL I Liaison Graph [50] Extracting the parts of a table-top vise assembly (18) ind C NS M NL C Connector-based Graph [89] Constructing the connector-based graph for a multi-plate clutch (27) ind C NS M NL C Visibility Graph [35] Moving a part through map of parts (6) puz C S M L I [37] Extracting a box from its associated cover (2) puz C S M L C [82] Removing parts of a model aircraft vertical tail (N/A) ind C S M L C Contact State Graph [46] Inserting an elastic tube into a rigid groove (2) ind C S M L C Roadmap-and Constraint-based [66] Removing a bolt from a pump (2) ind C NS M L I…”

“…having a distance smaller than a predefined threshold), as shown in Figure 11(b). Lee and Shin [50] presented a method for automatic determination of preferred assembly partial orders using liaison graph, in which the feasibility of an assembly operation is checked by calculating directional freedom of motion, manipulability, interference with neighboring parts, and accessibility of a part. Determining an assembly partial order is based on recursive extraction of preferred subassemblies and verification of their disassemblability.…”

<title>Planning and control of a microassembly process in a flexible microrobot-based desktop station</title>

“…There have been considerable research efforts in computer aided assembly planning, because it provides a systematic way to search for an optimal solution [6][7][8][9][10][11][12][13]. But there are some specific problems in an MMS, which are caused by the microscopic dimension of the parts to be assembled and by the limitations of station's work space.…”

“…the degree of parallelism, for the kth assembly plan; a e [0, 1] represents the weighting coefficient, so that a + a2 ÷ = The determination of the components of (28) and (30) is a rather difficult task that may affect the quality of assembly planning to a great extent. The operation cost representing an operation complexity is a function of interconnection type, such as attachment, sticking, force-fit, screw or welding [6]. Let us introduce the coefficient cck(O/O) of operation complexity, which corresponds to the interconnection between subassemblies O and 9 in the configuration ek.…”

A modularized contact-rule reasoning approach to the assembly sequence generation for product design

“…Tonshoff (1992) suggested that the data created from task-related, object-related, and organization rules could be transferred to assembly information and used for sequence planning. Lee and Shin (1993) assigned part details to nodes and established merge rules, such that the relations between the details and attributes of liaison graphs could be formed based on the rules. However, the method could be costly and time-consuming when applied to a large number of detailed components.…”