It is a common view that the assembly process heavily affects a product's final quality and cost. The continuously shortening product life cycle requires a faster response speed as well as a lower defect rate in assembly production. In this situation, assembly quality control is becoming one of the most demanding problems in the modern manufacturing environment. The main causes of assembly defects can be classified into four categories, i.e., improper design, defective part, variance in assembly system, and operator error. The first three categories have been studied for many decades. However, elements of the operator error have not been fully explored. In this paper, using a copier assembly as an example, the problem of assembly defects caused by mistakes of operators is investigated systematically. A novel defect-rate prediction model is derived from the study of two complexity factors, namely, the design-based assembly complexity factor and the process-based assembly complexity factor, which are defined according to the structure and production characteristics of the copier. Several case studies consistently demonstrate that the new prediction model is accurate and stable for evaluating the copier assembly quality. Moreover, another case study offered in this paper demonstrates that the prediction model can provide effective assistance in the improvement of assembly quality.

This paper addresses the problem of determining the instantaneous state of mobility and constraint of a general mechanism or assembly using screw theory. This problem has been attacked before by different methods with varying success. The method presented here is very simple and applies to many types of mechanisms, with one interesting exception. This exception also eludes the Griibler/Kutzbach criterion. The method presented here also improves the accuracy and scope of a method presented by Konkar and Cutkosky. Note to Practitioners - This paper permits an engineer to analyze the state of constraint of an assembly, mechanism, or fixture in order to see if it is properly constrained or whether it contains unwanted overconstraints. Overconstraint can lead to locked-in stress, incorrect tolerance analysis, and lack of robustness in a design. Very small changes in the sizes of parts or positions of locators in fixtures can have large effects on locations of parts, deformation, or internal stresses. Proper constraint is desirable because it provides for only one way to assemble parts. Operator dependency is eliminated, assembly is quick and repeatable, and special skin is not needed.

This paper is a review of flexibility issues in manufacturing. It covers strategies and approaches across the spectrum of product development, manufacturing, and distribution, showing different objectives of flexibility, such as design, manufacturing, and customization, along with approaches that have been suggested by researchers or implemented in industry.

This paper describes a prototype software system that implements a top-down approach to the concept stage of assembly design and analysis. The software implements a number of concepts and techniques introduced by the authors in previous publications. The study comprises a definition of a top-down approach to assembly design, development of several new assembly analysis and synthesis algorithms, and integration with existing algorithms. New assembly analysis and synthesis tools include: a graphical layout tool for defining an assembly level dimensional control plan, mathematical models of assembly features, assembly constraint analysis, variation propagation analysis in the presence or absence of adjustments, and an optimal control algorithm for assembly feature synthesis. Existing tools include assembly sequence constraint generator and assembly sequence generator-editor. A common assembly database is created that is used by the different analysis tools. Some of the functionalities are still evolving.

Uses assembly sequence analysis (ASA) to explore design for assembly (DFA), subassembly partitioning, and assembly sequence choice for two complex assemblies. Complex assemblies have very high parts-counts, a final assembly organized as an assembly of subassemblies, and offer limited redesign options. ASA addresses combinatorial aspects of complex assemblies that conventional DFA ignores: choice and partitioning of subassemblies, and assembly sequence choice. The paper describes criterion-based searches for favorable subassembly partitioning and assembly sequences that use genetic algorithm techniques to spread assembly move difficulty across entire final assembly sequences while satisfying all logical constraints imposed on the assembly sequence by part geometry. The measure of assembly move difficulty, a count of kinematic degrees of freedom secured during each final assembly step, is measured on an absolute scale. We find that ASA can pinpoint candidate DFA-related redesigns and can suggest assembly issues to designers. Logical assembly issues dominate quantitatively-characterized issues when selecting assembly sequence or subassembly partitioning. After logical issues are addressed, the sequence choice criterion defined here often duplicates choices made by experienced analysts. Finally, the sequence choice criterion favors in-line over branched final assembly lines.

Presents algorithms to propagate and control variation in mechanical assemblies using the state transition model approach. It exploits the modeling environment and uses concepts from control theory to model variation propagation and control during assembly. The assembly process is modeled as a multistage linear dynamic system. Two types of assemblies are addressed: Type-1 where the assembly process puts together parts at their pre-fabricated mating features and Type-2 where the process can incorporate in-process adjustments to redistribute variation. Algorithms are developed to determine and control variation in final assembly propagated through the combined effect of individual part variations and choice of assembly methods. Algorithms to propagate variation in the presence of adjustments are also presented. In-process adjustments in Type-2 assemblies are determined by the type of interface features between parts being assembled which are modeled as control inputs to the dynamic system. An optimal control problem is formulated to design these interfaces. Variation associated with final assembly dimensions, cost of making adjustments, and assembly sequence effects are included in the optimization procedure.

Presents a state transition model of assembly and concepts from control theory to model variation propagation and control during assembly. The assembly process is modeled as a multi-stage linear dynamic system. Two types of assemblies are addressed: Type-1 where the assembly puts together parts at their pre-fabricated mating features, and Type-2 where the assembly process can incorporate in-process adjustments to redistribute variation. The model builds on the concept of the datum flow chain. Algorithms are developed to determine and control variation in final assembly propagated through the combined effect of individual part variations and choice of assembly methods. An optimal control problem is formulated to develop a scientific approach to designing assembly features.

The phenomenon of vibration-assisted entrapment is analyzed in the context of a vibratory parts feeder - the SONY automated parts orienting system (APOS). Entrapment is treated as an energy minimization process. Mechanisms of energy increase and decrease are described. Conditions that lead to failure of entrapment are discussed for a planar part-relief system.

Applies simulated annealing (SA) to find the probable least cost assembly sequence for a mechanical product from among the thousands of feasible sequences. The sequences are represented by the assembly sequence network introduced by De Fazio and Whitney (1987). Candidate sequences are selected one at a time by the SA algorithm and their cost is estimated by designing a minimum unit cost concept assembly system using a dynamic programming algorithm. Alternate sequences are selected by perturbing the path through the sequence network in a way similar to a genetic algorithm. This method differs from that normally used to select least cost sequences, in which an engineer edits the thousands of sequences using judgment and experience, and then obtains the costs of a few that survive editing. Here, the search for a least cost sequence is done without prior editing. In the examples studied, there appear to be a large number of sequences that are technically quite different but of similar cost.>

This paper proposes an intelligent manufacturing system that can make decisions about the process in light of the uncertain outcome of these decisions and attempts to minimize the expected economic penalty resulting from those decisions. It uses robot weld bead grinding as an example of a process with significant process variation. A three tier hierarchical control system is proposed to plan an optimal sequence of grinding passes, dynamically simulate each pass, execute the planned sequence of controlled grinding passes, and modify the pass sequence as grinding continues. The top tier, described in this paper, plans the grinding sequence for each weld bead, and is implemented using stochastic dynamic programming, selecting the volumetric removal and feedspeed for each pass in order to optimize the satisfaction of the task requirements by the entire grinding sequence within the equipment, task, and process constraints. The resulting optimal policies have quite complex structures, showing foresight, anxiety, indifference, and aggressiveness, depending upon the situation.>

An integrated computer aid that is useful for assembly-line design and for concurrent design of mechanical products is described. Recognizing that early consideration of assembly sequence is important for producibility, quality control, flexibility, and market responsiveness, the authors have built an integrated set of user-interactive computer programs that generates all feasible assembly sequences for a product and then aids the user in judging their value based on various criteria. The programs use a disassembly analysis for generating sequences and provide online visual aids during generation and evaluation. During evaluation, matters such as avoiding difficult assembly states or moves, stability, fixturing, orientation, refixturing and reorientation count, and inclusion of favorable states are considered to highlight desirable or undesirable sequences. The designer edits the set of sequences according to these criteria, leading to an informed sequence choice or to needed design refinement. The interactive programs provide a rapid means for sequence selection, encouraging their use during early design.>

The authors discuss computer aids for assembly sequence choice and design during product design. Implementations include adapting A. Bourjault's (1984) method for finding physically possible assembly sequences, using a solid modeler during this step, representing sequences in a state transition space, and reducing the sequence count to a small set of desirable ones by computer-interactive editing of the state transition space.>

Bourjault has presented a method which will generate all valid assembly sequences for the set of parts that constitute an assembly. A modification of Bourjault's method is presented which makes practical the application of this technique of assembly analysis to assemblies with greatly increased part count. The salient difference between the two methods is in the form and number of the questions whose answers yield the relations that allow algorithmic generation of assembly sequences. Bourjault's method requires 2l^{2} questions plus an often-large number of subsequent questions whose existence depends on answers to part of the former question set; all have yes or no answers. (Here l is the number of relations between parts.) The modified method requires 2l questions that are answered in a precedence-logical form; the questions are similar to those asked by an engineer contemplating assembly of a set of parts. Applications and techniques for use are presented, and examples are shown for l as great as 18.

Robotics is a basic research field with a growing mandate to meet society's needs. I argue in this paper that robotics needs a new research agenda and set of guiding assumptions before it can fill both the basic and applied roles. An inadequate dependence on mimicking humans must be replaced by an outward looking, interdisciplinary effort that recognizes the importance of understanding the basic processes robots will do. Robotics restructured this way can overcome its focus on devices and become the technology that forces basic knowledge advances in processes, product design, manufacturing technology, and manufacturing organization.

Bazerghi et al. present an "exact kinematic model of the PUMA." It appears to be a model of the PUMA's internal software model rather than a model of the PUMA itself. Other sources of information about nominal parameter values in the PUMA's internal software model are compared with actual parameter values obtained from calibrations of two PUMA 560 manipulators. It has been shown that the actual PUMA and its internal model disagree about tool tip position, typically by several millimeters.

This presentation combines historical lineage, assessment of the state of the art, and representative videotapes on robot force control. The difference between continuous and logic branching strategies is described. The development of various impedance strategies and hybrid methods is traced and compared. The problem of stability is discussed and remedies are related to higher strategy issues.

CADSYS is an interactive computer-aided decision support system which interfaces a designer with a mathematical, rather than graphic, model of a design problem. Multiple design goals can be imposed in the form of equality and inequality constraints, whose levels can be manipulated. A numerical search is used to find designs which meet the constraints. CADSYS has now been applied to two real and current design problems: the hydraulic system of a novel inertial navigation unit and the financial, technical, and logistical problem of liquified natural gas tanker design and operation. Both of these studies returned new understanding of the problems themselves and of the capabilities of CADSYS.

This paper presents a first implementation of a computer-aided design concept intended to confront issues normally faced by designers of complex engineering systems. Computer graphics play no essential role in this concept. Rather, the concept is aimed at organizing the analysis and search activities of a designer as he attempts to find values for the independent design parameters at his disposal so that specific performance goals will be met. The framework for the implementation is an interactive computer system written in PL/I, comprising a problem-independent command interpreter and a completely general problem-dependent analytical statement relating the design parameters and the performance measures. The designer can change the performance specifications (and to a limited degree the analytical problem statement itself) and can issue commands directing searches for solutions (values for the independent parameters) that satisfy, or come closest to satisfying, the requirements. If the requirements conflict and admit of no solution, he can relax them in an organized fashion. If they are slack and admit many overdesigned solutions, he can search for ways to tighten them or to balance the achievement of them. A manager can assign several designers to use computer-aided design system (CADSYS) simultaneously to work on related parts of a large problem. Changes or results by one designer are immediately felt by the others and by the manager. CADSYS thus has some of the properties of a management information system.