An Artificial Intelligent based System to Automate Decision Making in Assembly Solution Design

Manssouri

2022

JERA

Over the last decades, there has been growing pressure on industrial companies to offer to their costumers products with high quality, in the minimum deadlines and with reasonable prices. Since the design phase plays a key role to achieve these difficult goals, many traditional, DFX (Design For X) and integrated approaches have been proposed. However, many limits are still present. Thus, the main objectives of this work were first to identify these limits and then to overcome them by proposing and developing an automated framework for integrated product design. In this work, we automated the integrated DFMMA (Design For Materials, Manufacturing and Assembly) approach by developing an architecture composed of four levels, namely: the Common Information Modeling Level, the Selection Systems Level, the Inference and Computation Level and finally the Application Level. The proposed automated system is based on ontologies, on the CBR (Cases Based Reasoning) and the RBR (Rules Based Reasoning). The first main result obtained throughout the contributions consists on the integration of Manufacturing process selection, Assembly solution selection and materials selection in one integrated design approach. The second main result obtained consists on the exploitation of all the previous design studies developed by the design team and the ability to reuse the designers experience throughout the case based reasoning used in the proposed architecture. Another important result consists on the formalization and the automation of the execution of the design rules and the ability to infer new results and to check inconsistencies in the developed product using the data and information modeled in the ontological model and throughout the Cases Based Reasoning that we have incorporated in the developed approach. In this way, the redundancy in work and the difficulties faced in case of having a high number of design alternatives are avoided. Consequently, the product quality increases and wastes of time and money decrease. Finally, to validate the functioning and the efficacy of the proposed DFMMA system, an application on the design of a complex mechanical product is developed in the end of the work.

Section: The Selection Systems Levelmentioning

confidence: 99%

An Artificial - Intelligent - Based System to Automate the Design of Complex Mechanical Products

Manssouri

2022

JERA

“…The consistency of part A in multiple variants might enable the standardization of other processes. The knowledge-based methodologies of Das and Swain [40] and Chaimae et al [165] have considered similar joining requirements for new designs, but they lack concrete product variety considerations.…”

Section: Discussionmentioning

confidence: 99%

“…However, they still needed an MCDM module to analyze the feasibility and fitness of solutions. Chaimae et al [165] also presented an ontology-based approach that uses case-and rule-based reasoning (see Section 2.3). On a more abstract methodology selection level, Hoefer and Frank [166] used ML to select manufacturing processes.…”

Section: Technology and Parameter Selectionmentioning

confidence: 99%

“…Recent works using CBR include the knowledge-based frameworks in JTS methodologies (e.g., [40,165]).…”

Section: Artificial Intelligencementioning

confidence: 99%

“…However, it requires extensive development and maintenance effort, mainly due to the vast solution space and many distinct joining scenarios. Furthermore, liaison-based methodologies include the CBR approach of Das and Swain [40] and the ontology approach of Chaimae et al [165].…”

Section: Frameworkmentioning

confidence: 99%

See 2 more Smart Citations

Automated joining element design for high product variety in the manufacturing industry

Eggink

Product variety and the manufacturing complexity that it induces are continuously increasing. This poses a challenge in the product development process and, consequently, the design of joints. Joining elements define the manner in which a permanent joint is created between parts. Joining element design is an ambiguous manual task with limited automation solutions. Thus, it can lead to long, iterative, error-prone development trajectories that may result in costly rework. Hence, automation solutions for joining element design must be intelligent. However, simply ensuring the intelligent automation of joining element design is insufficient. Modular design, through the approaches of modularization and commonalization, enables manufacturers to cope with the complexity induced by product variety. Unfortunately, modular design approaches have not yet considered joining elements. Hence, this dissertation study sought to answer the following research question: "How can joining element design be automated for high-variety products?"This dissertation presents a framework for automating joining element design. The framework is structured into smaller design problems, which will guide designers through the process of designing joining elements. This structuring will also enable designers to evaluate and assess artificial intelligence (AI) for each design problem. Moreover, the design problems will enable designers to identify unused AI techniques, such as machine learning.These techniques were conceptualized and several were implemented for validation in this study. A market-validated database with automotive Body-in-White structures was used. The validation included the use of decision trees to predict joining technologies and the number of joining elements. Both prediction tasks seem promising to implement in early design phases due to their simplicity.Next, this study validated two approaches for predicting joining locations. The first was a straight-forward evolutionary algorithm for distributing spot welds over contact regions. However, this algorithm's performance fell off quickly for nontrivial design problems with high solution spaces.The second approach was to use convolutional neural networks to predict spot weld locations, which provided robust and promising results. This study first explored a voxel-based regression and classification task by drawing locations in a grid-like structure. Classification with a segmentation approach produced more robust results due to the spatial dependencies of classes. Supervised machine learning enabled the consideration of knowledge of successful designs in new design problems. This concept was subsequently enriched with nongeometric data using a branding approach. However, these multimodal machine learning models did not improve the joining locations. Furthermore, the models could not extract the i ii

Towards the definition of assembly-oriented modular product architectures: a systematic review

Monetti,

Maffei

2023

Res Eng Design

The success of a product in the market is largely defined by the quality of design decisions made during the early stages of development. The product design requires designers to balance multiple objectives such as functionality, cost, and user satisfaction, while addressing the challenges posed by increasing product variants and customization demands. To tackle these challenges, one approach is to structure a comprehensive model that incorporates design for assembly (DFA) guidelines during the formulation of product architecture in the conceptual phase of development. While numerous strategies have been proposed in the literature, information is often scattered, making it difficult for readers to gain a comprehensive understanding of the topic. This paper systematically reviews the role and impact of DFA in product development, consolidating and presenting the information coherently. The review provides an overview of the methods developed, along with their potential benefits and limitations. A common framework is identified that defines the structure of the models, helping designers integrate assembly consideration into their design processes, thus reducing assembly time, cost, and complexity. The framework describes the operational setting, including the domain and context in which models operate, and offers a classification of possible methods and desired outputs. Additionally, the review identifies the industry in which case studies have been most frequently presented, and the software used to facilitate the process. By connecting with such a framework, future models can be created following a structured approach, and existing models can be classified and upgraded accordingly.