Assembly Sequence Planning Research Articles

Multi-objective multi-verse optimiser for integrated two-sided assembly sequence planning and line balancing

Multi-objective multi-verse optimiser for integrated two-sided assembly sequence planning and line balancing

Assembly Sequence Planning: A Review

Background: Assembly Sequence Planning (ASP) is an important stage in the process of product design and manufacturing. A reasonable assembly sequence can reduce the complexity of assembly operations, reduce the number of tools and fixtures, improve assembly efficiency, reduce assembly time, and reduce assembly costs. Methods: The purpose of this paper is to review the objective function, constraint conditions, and the solving methods of the ASP. Conclusion: The research methods can be roughly divided into three categories: graph-based method, knowledge-based method, and artificial intelligence algorithm. The advantages and disadvantages of the three methods are compared. Finally, future research is discussed, which could provide a reference for the ASP.

Dynamic multi-constrained assembly sequence planning of large space structures considering structural vibration

The assembly sequence planning problem of large space structure (LSS) is investigated in this paper, from the view of structural vibration. With the goal of reducing structural vibration and improving structural stability during assembly, the whole planning problem is decomposed into multiple problems of optimal assembly positions at each operation. Adopting the plate-type LSS and the truss-type LSS as specific objectives, a dynamic multi-constrained assembly sequence planning algorithm is then proposed based on structural dynamics modelling and modified genetic algorithm (GA). Selecting the structural frequencies to quantitatively describe structural stability of LSS, the algorithm is implemented through a designed procedure including an external loop and an internal loop. The dynamic assembly constraints, including geometric constraint, dynamics constraint, non-repeatability, realizability and continuity, are handled in the internal loop through assembly information delivered from the external loop. Numerical simulations of the two types of LSS are finally provided to testify the feasibility of the proposed sequence planning algorithm. The results show that the obtained sequences can effectively make LSS characterized by weak structural vibration during assembly.

Assembly Sequence Planning with Deformable Linear Objects in the Smart Factory: Dilemmas and Injections

Smart factory consists of extensive automation (e.g., robotic actors) and minor human interventions. Additional characteristics are manufacturing a variety of products rapidly and accurately. Many of the products contain rigid objects assembled with deformable ones, especially deformable linear objects (DLO) such as wires. Deformable object behavior results in modifications of objects' shape and position thus influencing assembly processes. Assembly sequence planning (ASP) is an intractable issue in production management. ASP in a robotized environment such as the smart factory with objects deformations is even more challenging. Deformable objects assembly by robotic actors increase uncertainty and influence factory performances. The theory of constraints (TOC) suggests tools for improving factory performances by concentrating on production flow. The literature is lacking TOC-based tools investigating smart factories and Industry 4.0 aspects. One of TOC thinking process tools is a conflict resolution diagram. This diagram describes a dilemma influencing achieving desired goal and aids in identifying potential solutions (called "injections"). This paper describes ASP dilemma in smart factories regarding robotic manipulation with deformable objects. The dilemma between generic, unified automatic assembly on one hand and customized, designated assembly, on the other hand, is elaborated. An entity called "interaction" is introduced as an injection to the smart factory ASP dilemma. Interactions are defined based on physical, geometric, and manufacturing attributes. Defining interactions categories is an important step toward comprehensive automation of assemblies containing deformable objects. By so interactions are promising direction from both research and industrial perspectives for robotic ASP with deformable objects in smart factories.

A Survey of Space Robotic Technologies for On-Orbit Assembly

The construction of large structures is one of the main development trends of the space exploration in the future, such as large space stations, large space solar power stations, and large space telescopes. It is one of important development tendency, which aims to make full use of space robots to assemble space structures autonomously in the aerospace industry. Considering that on-orbit assembly is an effective method to solve the problem of construction of large-scale spatial structures, it is necessary to motivate and facilitate the research of space robotics technologies for on-orbit assembly. Therefore, in this paper, the development status of space robot technology and the relevant space robot on-orbit assembly technology in recent decades are summarized. First, based on the space robot motion planning and assembly sequence planning, the development of space robot planning algorithms is introduced. For space robot assembly task, the space robot assembly method is summarized. From the control point of view, how to solve the vibration suppression and compliant assembly of on-orbit assembly is reviewed, which provides a reference for the autonomous intelligent assembly of space robots for large-scale structures in space. In order to simulate the space assembly scene on the ground, this paper introduces the development of ground verification experiments and provides ideas for the effective verification of space on-orbit assembly technology. In summary, though some of these problems have been satisfactorily solved in the past research, further research is still necessary in the future. Finally, it looks forward to the future research direction of space machine on-orbit assembly.

CAD-based AND/OR Graph Generation Algorithms in (Dis)assembly Sequence Planning of Complex Products

Several studies in production engineering and operations research focused on disassembly and assembly planning for improving the profitability of remanufacturing. The representation of feasible (dis)assembly sequences by analyzing geometrical and technical precedence constraints is an essential necessity of (dis)assembly planning. An AND/OR graph (AOG) is commonly used to represent a product’s feasible (dis)assembly sequences, including alternative subassemblies and parallel operations. In recent years, researchers studied the automatic extraction of geometrical precedence constraints by collision analysis within 3D models. However, since most of the existing approaches focused on collision analysis to identify complex precedence constraints, the generation of AOGs for complex products from collision analysis results remains too inefficient for industrial use cases. In this study, a Computer-Aided Design (CAD) interface from previous studies is used to extract liaison and moving wedge information from 3D models. A top-down and a bottom-up approach for generating complete AOGs from the extracted data are introduced. The approaches’ computing performances are analyzed on sample test cases and CAD models. The bottom-up approach performed better for the given samples. It has been found that the amount of moving wedge constraints has a strong effect on computing performance. It is a strong indicator to estimate the complexity of products under examination. The exponential behavior of the needed computing resources can be estimated beforehand. For complex products, graph simplification or alternative graph representations with less information richness should be considered. The results contribute to automated (dis)assembly sequence planning from complex products and, consequently, increasing remanufacturing profitability.

Automatic identification of the assembly base component for robotic manufacturing

The base component of an assembly is the part placed on the worktable, and typically the part from which the assembly sequence starts. Automatically determining a base component suitable for robotic assembly is challenging since both assembly requirements and part handling requirements must be considered. However, automatically determining the base component is an important preliminary stage for automatic assembly sequence planning, which can considerably increase process efficiency. Prior work on automatic base component identification concentrated on the assembly requirements without specific regard to the requirements of robotic assembly. The current work integrates assembly requirements with robotic handling requirements. A two-stage algorithm for identifying a base component suitable for robotic assembly is presented. A geometric stage that addresses assembly requirements and is based on the geometric parameters of the parts extracted from a CAD model. A caging stage that addresses robotic assembly requirements assessing whether the parts can be caged by a standard vice. The algorithm was implemented with SolidWorks API and tested with ten assemblies with 2-8 parts each. The selections made by the algorithm were compared to selections made by process experts. The agreement between the experts and the algorithm was 77% and 90% for the geometric and caging stages respectively. The high agreement in the caging stage selection validates the suitability of the algorithm as a preliminary stage for automatic robotic assembly sequence determination.

A sub-assembly division method based on community detection algorithm

ABSTRACT Assembly sequence planning is always a great challenge in complex product assembly. Subassembly division is a way that can validly decrease the complexity of assembly sequence planning problem. Above all, assemblies comprising parts with different connection relationships can be modelled as the weighted graph in which the weighted adjacent list is used to accomplish calculation of the subassembly division. A weighted adjacency list generation method is presented for generating the list quickly. Furthermore, according to the requirements of subassembly division and the connection relationships of actual parts, an optimized community detection algorithm based on Fast Newman algorithm is proposed. The weight assignment of different connection relationships and deviation parameters are introduced to improve the subassembly division results. Finally, the optimized algorithm is applied in the subassembly division process of a practical pumping unit, with the influence of the weight assignment and deviation parameters being analyzed in detail. The results show that the proposed algorithm has a good performance in terms of balance, stability and independence.

Assembly Sequence Planning Using Artificial Neural Networks for Mechanical Parts Based on Selected Criteria

The proposed model of the neural network describes the task of planning the assembly sequence on the basis of predicting the optimal assembly time of mechanical parts. In the proposed neural approach, the k-means clustering algorithm is used. In order to find the most effective network, 10,000 network models were made using various training methods, including the steepest descent method, the conjugate gradients method, and Broyden–Fletcher–Goldfarb–Shanno algorithm. Changes to network parameters also included the following activation functions: linear, logistic, tanh, exponential, and sine. The simulation results suggest that the neural predictor would be used as a predictor for the assembly sequence planning system. This paper discusses a new modeling scheme known as artificial neural networks, taking into account selected criteria for the evaluation of assembly sequences based on data that can be automatically downloaded from CAx systems.

A brief review of simulated Kalman Filter Algorithm – variants and applications

Simulated Kalman Filter (SKF) solves optimization problems by finding the estimate of the optimum solution. As a multi-agent algorithm, every agent in the population acts as a Kalman filter by using a standard Kalman filter framework, which includes a simulated measurement process and a best-so-far solution as a reference. This paper presented an overview of the research progress in SKF from the day it was introduced until the present day, discussing the progress, improvements, modifications, and applications of SKF. The fundamental and standard algorithm were first introduced. Then, the work on the algorithm improvements was surveyed. Finally, the remaining unresolved problems and some directions of SKF research were discussed. We reviewed 57 SKF papers. 16 of them on fundamental improvements, 9 on extension of the algorithm to discrete problems and 25 on their applications. Researchers have worked on ideas to improve exploration capability to prevent premature convergence by trying prediction operators, opposition-based learning, and different iteration strategies. There were also attempts to hybridize SKF with other famous algorithms such as Particle Swarm Optimization (PSO), Gravitational Search Algorithm (GSA), and Sine Cosine Algorithm (SCA) to improve its performance. Lastly, a single-agent variant of SKF and a multi-objective SKF were introduced. SKF algorithms and its variants have been implemented in at least nine areas of applications: drill path optimization, airport gate allocation problem (AGAP), assembly sequence planning (ASP), system identification, feature selection, image template matching, controller tuning, wireless sensor network, and engineering design problem. The literature reviewed solely depended on the keyword search that contained the terms simulated Kalman filter from December 2015 to the present date. This is the first review paper on SKF. It is hoped that this survey would be beneficial for the researchers of this area and attracting interest towards the algorithm.

A modified cut-set method for mechanical subassembly identification

PurposeThe purpose of this paper is to perform a comparative assessment on working of the existed subassembly identification (SI) methods, which are widely practiced during the product development stage and to propose an improved method for solving the SI problem in assembly sequence planning (ASP).Design/methodology/approachThe cut-set method is found as a suitable method among various knowledge-based methods such as the theory of loops, theory of connectors and theory of clusters for the workability enhancement to meet the current requirements. Necessary product information is represented in the matrix format by replacing the traditional AND/OR graphs and the advanced predicates are included in the evaluation criteria.FindingsThe prominent methods in SI are followed a few of the predicates to avoid complexity in solution generation. The predicate consideration is found as the most influencing factor in eliminating the infeasible part combinations at SI. However, the quality of identified subassemblies without advanced predicates is not influencing the solution generation phase but practical applicability is affecting adversely.Originality/valueThe capability of performing SI by the cut-set method is improved to deal with the complex assembly configurations. The improved method is tested by applying on different assembly configurations and the effectiveness is compared with other existent methods of ASP along with the conventional method.

Measuring the effect of automatically authored video aid on assembly time for procedural knowledge transfer among operators in adaptive assembly stations

ABSTRACT Can automatically authored videos of industrial operators help other operators to learn procedural tasks? This question is relevant to the advent of the industrial internet of things (IIoT) and Industry 4.0, where smart machines can help human operators rather than replacing them in order to benefit from the best of humans and machines. This study considers an industrial ecosystem where procedural knowledge (PK) is quickly and effectively transferred from one operator to another. Assembly tasks are procedural in nature and present a certain complexity that still does not allow machines and their sensors to capture all the details of the operations. Especially if the assembly operation is adaptive and not fixed in terms of assembly sequence plan. In order to help the operators, videos of other operators executing the complex procedural tasks can be automatically recorded and authored from machines. This study shows by means of statistical design and analysis of experiments that expert aid can reduce the assembly time of an untrained operator, whereas automatically authored video aids can transfer PK but producing an opposite effect on the assembly time. Therefore, hybrid training methods are still necessary and trade-offs have to be considered. Managerial insights from the results suggest an unneglectable impact of the choice to digitise industrial operations too early. The experimental studies presented can act as guidelines for the correct statistical testing of innovative solutions in industry.

An integrated job shop scheduling and assembly sequence planning approach for discrete manufacturing

This paper proposes an integrated job shop scheduling and assembly sequence planning (IJSSASP) approach for discrete manufacturing, enabling the part processing sequence and assembly sequence to be optimized simultaneously. The optimization objectives are to minimize the total production completion time and the total inventory time of parts during production. The interaction effects between the job shop schedule and the assembly sequence plan in discrete manufacturing are analyzed, and the mathematical models including the objective functions and the constraints are established for IJSSASP. Based on the above, a non-dominated sorting genetic algorithm-II (NSGA-Ⅱ) with a hybrid chromosome coding mechanism is applied to solve the IJSSASP problem. Through the case studies and comparison tests for different scale problems, the proposed IJSSASP approach is verified to be able to improve the production efficiency and save the manufacturing cost of the discrete manufacturing enterprise more effectively.

A decision-making method for assembly sequence planning with dynamic resources

With the advent of mass customisation, solving the assembly sequence planning (ASP) problem not only involves a non-convex optimisation problem that is hard to solve but also requires a high-speed response to the changes of assembly resources. This paper proposes a deep reinforcement learning (DRL) approach for the ASP problem, aiming at promoting the response speed by exploiting the reusability and expandability of past decision-making experiences. First, the connector-based ASP problem is described in a matrix manner, and its objective function is set to minimise assembly cost under the precedence constraints. Secondly, an instance generation algorithm is developed for policy training, and a mask algorithm is adopted to screen out impracticable assembly operations in each decision-making step. Then, the Monte Carlo sampling method is used to evaluate the ASP policy. The policy is learned from an actor–criticbased DRL algorithm, which contains two networks, policy network and evaluation network. Next, the network structures are introduced and they are trained by a mini-batch algorithm. Finally, four cases are studied to validate this method, and the results are discussed. It is demonstrated that the proposed method can solve the ASP problem accurately and efficiently in the environment with dynamic resource changes.

Assembly sequence and path planning for monotone and nonmonotone assemblies with rigid and flexible parts

• SPP-Flex method for assembly sequence and path planning of rigid and flexible parts. • A new greedy assembly sequence planning heuristic minimizes assembly stress on parts. • A new sampling-based path planner (BXXT) is proposed for path planning of parts. • Assembling flexible parts is simulated and analyzed via the Abaque FEA software. • Four variants of SPP-Flex are proposed to solve monotone and nonmonotone products. The Assembly Sequence and Path Planning (ASPP) problem deals with finding a proper sequence of parts to be assembled into a finished product and short assembly paths for each part. The problem is a combination of Assembly Sequence Planning (ASP) and Assembly Path Planning (APP) subproblems, which are both NP-complete and therefore intractable at large sizes. Nearly in all works on ASPP, it is assumed that planning is monotone (i.e., parts are moved only once, without considering intermediate placements) and each part is completely rigid. These are simplifying, yet limiting assumptions, since most assembled products like ships, aircraft, and automobiles are composed of rigid and flexible parts, and the generation of assembly sequence and path plans for most real-world products requires intermediate placement of parts to be taken into account. None of the existing works in the literature, however, have handled nonmonotone ASPP problems for rigid and flexible parts, and this issue remains largely untouched. In this paper, we present a new method called SPP-Flex for solving monotone and nonmonotone ASPP for rigid and flexible parts. SPP-Flex first utilizes a Directional Assembly Stress Matrix (DASM) for describing interference relations between all pairs of parts and the amounts of compressive stresses needed for assembling flexible parts and then obtains an initial tentative assembly sequence using a simple new greedy heuristic. Next, short assembly paths are iteratively computed and planned from initial to goal configurations of all parts using a novel sampling-based path planner called BXXT. If finding a free path for an active part fails due to obstruction of a previously assembled part, then such a part is identified, relocated, and its path replanned until the active part is moved to its final position. In case of failure again, if the part is flexible, through finite element analysis, it is determined if the part can still be assembled by undergoing elastic deformation. To evaluate the performance of the SPP-Flex and its components, two new products were designed and solved by four combinations of ASP and APP methods 20 times each, and the means and standard deviations of five performance criteria (total path length, total number of generated nodes and edges in the search tree, total number of collision (interference) checks, and total runtime) were calculated. Analysis of the computational results showed that the proposed greedy heuristic sequence planner together with the BXXT path planner/replanner outperformed other variations with at most 4.6% average gap in path length and 2.1% average gap in runtime compared to the best-found solution in all runs.

Towards online reinforced learning of assembly sequence planning with interactive guidance systems for industry 4.0 adaptive manufacturing

Literature shows that reinforcement learning (RL) and the well-known optimization algorithms derived from it have been applied to assembly sequence planning (ASP); however, the way this is done, as an offline process, ends up generating optimization methods that are not exploiting the full potential of RL. Today’s assembly lines need to be adaptive to changes, resilient to errors and attentive to the operators’ skills and needs. If all of these aspects need to evolve towards a new paradigm, called Industry 4.0, the way RL is applied to ASP needs to change as well: the RL phase has to be part of the assembly execution phase and be optimized with time and several repetitions of the process. This article presents an agile exploratory experiment in ASP to prove the effectiveness of RL techniques to execute ASP as an adaptive, online and experience-driven optimization process, directly at assembly time. The human-assembly interaction is modelled through the input-outputs of an assembly guidance system built as an assembly digital twin. Experimental assemblies are executed without pre-established assembly sequence plans and adapted to the operators’ needs. The experiments show that precedence and transition matrices for an assembly can be generated from the statistical knowledge of several different assembly executions. When the frequency of a given subassembly reinforces its importance, statistical results obtained from the experiments prove that online RL applications are not only possible but also effective for learning, teaching, executing and improving assembly tasks at the same time. This article paves the way towards the application of online RL algorithms to ASP.

Optimization of Assembly Sequence Planning of Turbine Low-pressure Rotor Blades Based on the Improved Simulated Annealing Algorithm

In view of the NP-hard characteristics of assembly sequence planning problem and the low efficiency of solving the relatively optimal assembly sequence, this paper adopts meta-heuristic algorithms to solve the rotor blades assembly sequence optimization problem, Taking solution accuracy, solution efficiency, and solution robustness as assessment criteria, the effectiveness of solving the rotor blades assembly sequence with several meta-heuristic algorithms is compared and analyzed, the comparison results show that the simulated annealing algorithm has the best solution accuracy, efficiency and robustness for the assembly sequence optimization problem of rotor blades. In order to solve some problems during the solving process of simulated annealing, one improvement is added into it, by adding the reheating process to the SA, the probability that the SA algorithm accepts inferior solutions is increased, and the ability of SA jumping out of the local optimal solution is increased, and the global search capability of the SA algorithm is further enhanced. Finally, the effectiveness of the improved simulated annealing algorithm is verified in this paper by the assembly of steam turbine rotor blades.

The Extraction of Product Assembly Feature Information for Intelligent Assembly Sequence Planning

The assembly feature information contained in the CAD model of the product is an important basis for the generation of intelligent assembly sequence. Taking the assembly model of the product as the research object, the research on the acquisition of assembly feature information of the product by intelligent assembly sequence planning is proposed in this paper. Microsoft Visual Studio 2005 and SolidWorks API are used as implementation means to access the internal data of the product CAD model and realize the extraction of the hierarchical information, mating relation and mating line and surface assembly features of the assembly model. An application example of product assembly feature information extraction for the locking device of an elevator landing door shows that the established method of acquiring product assembly feature information is effective and provides a theoretical and technical basis for the establishment of product design and assembly sequence planning knowledge base.

Ant colony optimization for assembly sequence planning based on parameters optimization

As an important part of product design and manufacturing, assembly sequence planning (ASP) has a considerable impact on product quality and manufacturing costs. ASP is a typical NP-complete problem that requires effective methods to find the optimal or near-optimal assembly sequence. First, multiple assembly constraints and rules are incorporated into an assembly model. The assembly constraints and rules guarantee to obtain a reasonable assembly sequence. Second, an algorithm called SOS-ACO that combines symbiotic organisms search (SOS) and ant colony optimization (ACO) is proposed to calculate the optimal or near-optimal assembly sequence. Several of the ACO parameter values are given, and the remaining ones are adaptively optimized by SOS. Thus, the complexity of ACO parameter assignment is greatly reduced. Compared with the ACO algorithm, the hybrid SOS-ACO algorithm finds optimal or near-optimal assembly sequences in fewer iterations. SOS-ACO is also robust in identifying the best assembly sequence in nearly every experiment. Lastly, the performance of SOS-ACO when the given ACO parameters are changed is analyzed through experiments. Experimental results reveal that SOS-ACO has good adaptive capability to various values of given parameters and can achieve competitive solutions.

Cyber-physical assembly system-based optimization for robotic assembly sequence planning

Robotic assembly plays a principal role in intelligent manufacturing and Industry 4.0. Well-informed coordination between the robotic arms and the control modules is of paramount importance in the design and planning of automated systems. Given the time-extensive nature of assembly sequence planning and the need for labor-intensive coding and coordination, knowledge-based automated systems are much-needed tools to help facilitate the development of intelligent manufacturing systems. This research puts forward a Cyber-Physical Assembly System (CPAS)-based metaheuristic for fully-automated assembly sequence planning taking into account the physical characteristics of robotic arms. Illustrative case examples are provided to evaluate the practicability of the proposed approach. It is shown that the proposed approach generates better assembly sequences compared to the traditional methods and that the single-arms robots can execute the resulting assembly plans effectively. The numerical analysis confirmed that the CPAS is more efficient than the existing interactive method most commonly applied for assembly sequence planning. Overall, the proposed framework sets the stage for more effective and fully automated planner-robot interaction in the assembly sequence planning context.