Guiovanni Jules

A holonic systems approach to the formation of manufacturing networks

This paper attempts to model the formation of a manufacturing network based on the holonic manufacturing concept. An industrial case study involving a large network of Small and Medium-sized Enterprises (SMEs) has been carried out to investigate the processes underpinning the network formation and job allocation activities. An algorithm using agent-based (Contract Net Protocol) features has been developed to validate the holonic model with real-life industrial data. The simulation software was calibrated so that the model responds with a lead time value close to that obtained in the real life network by altering the network capacity. The results proved that the software and the model possess a good potential, as a benchmark, for the comparison of optimised and new experimental models of a manufacturing network based on holonic system.

Challenges and trends in the allocation of the workforce in manufacturing shop floors

This paper presents a critical and focused review of definitions, challenges and developments of the workforce allocation system in manufacturing shop floors with a particular emphasis on response to changes and disturbances. The continuously changing requirements in production environments have been the driving force behind several stages of evolution experienced by management of workforce allocation. The paper includes an in-depth examination of the phenomena of changes and disturbances in the production shop floor. This is followed by an assessment of the readiness of workforce allocation systems to respond to such events with respect to both workforce perception and system architecture. Recent developments of the workforce allocation models suggest that the use of intelligent and distributed systems and techniques offer solutions that are responsive, flexible, and multi-criteria. To this end, an overview of a recent study in the application of the principles of holonic manufacturing systems, as an example of intelligent system approach to develop a flexible workforce allocation model, is presented.

Holonic Ontology and Interaction Protocol for Manufacturing Network Organization

The aim of this paper is to form the best collaborative networked organization of small manufacturers to fulfill a production plan. Manufacturing networks can be formed using two approaches. The conventional approach is the formation of networks by the grouping of the best manufacturers based on multiple criteria. The proposed approach is the formation of all feasible networks followed by the grouping of the best networks based on multiple criteria. The hypothesis of this paper is that the better group of manufacturing networks may not be the result of the selection of the best available manufacturers to form each network, based on the same criteria. The research methodology employed involves techniques such as interaction protocol development, ontology engineering, and principles from concepts such as the contract net protocol, the product-resource-order-staff-architecture, the auctioning process, agent-based modeling, and rule-based programming. Results from the proposed approach were compared to the conventional approach on multiple criteria namely lead time, quality, cost, and delivery reliability. On the basis of the results, the hypothesis was not rejected.

Holonic Goal-Driven Scheduling Model for Manufacturing Networks

This study is about the generation of prospective job schedules by manufacturing networks prior to selection of the best manufacturing network for the job. The primary objectives of the manufacturing network selection process were to find networks with the best job lead time (determined by the scheduling process), then those with the best job quality, and finally those with the best job cost. In this paper, forward scheduling of jobs was combined with the popular priority rules namely FCFS, SPT, DDATE and SLACK, and compared with instances when resource slack time reduction algorithm was used. A Computer-Aided-Software-Engineering tool was created to provide the infrastructure that inputted relevant scenarios to the scheduling algorithms and outputted the results in terms of the indicators of lead time, quality, cost and resource slack time. The tool is an implementation of the distributed and holonic manufacturing model and was used to develop the decision making algorithms. These algorithms are presented in this paper. Results show improvements in terms of aforementioned indicators. In the case of SPT priority rule, resource slack time improvements were encouraging.

Supplier selection using modified Fuzzy AHP-based approach in manufacturing networks

Due to the economy recession all around the world, the performance of manufacturing industry has been significantly affected. The question of how to generate fast response to customers, maintain the high quality, accurate delivery rate and low risk become the crucial issue of keeping a company competitive. Supplier selection, one of the key factors for a successful production process has attracted more attention in the recent decade. This research aims to understand the supplier selection procedure in Small to Medium-sized Enterprises (SMEs) manufacturing networks. Statistical method Kernel density estimation is integrated with Fuzzy AHP for the first time to improve the accuracy for the evaluation of suppliers against quantitative criteria. Data and information are collected from a central company which coordinates manufacturing networks of SMEs as an industrial case study. As a result, the proposed statistical analysis method for supplier evaluation with respect to quantitative criteria could successfully assist decision makers and reduce the potential bias from a visual aspect in the decision making process.

Ontological Extension of PROSA for Manufacturing Network Formation

The scope of the research is the inter-enterprise communication which is required for the formation of manufacturing network. The overall aim of the research is to develop a simulation platform for the formation of manufacturing networks, scheduling of jobs in the networks and an intelligent assistance system for mending the networks in the event of disturbances. Ontology is normally built from a foundational ontology. But in this paper, we use a reference architecture. The Product-Resource-Order-Staff Architecture PROSA is used as the basis for ontology development. This paper presents the development of an ontology used in the formation of manufacturing networks. The ontology is developed to support the entities, their interaction protocol and the knowledge emerging during the auctioning process in the manufacturing domain. The ontology is also used to enable an illustrated investigation of the interactions during the auctioning process. The methodologies involve ontology editing, object-oriented programming and the visualization of the objects and interactions during the network formation and scheduling process.

Holonic-based flexible allocation of workforce in labourintensive manufacturing shop floors

Assembly-based operations, in one-off and batch production, are mostly served by labour-intensive industries. The throughput performances of those industries are highly dependent on the performance of the labour on the job and the availability of the workers. The most significant threat to those industries is the sudden unavailability of skilled labour. Through a flexible workforce allocation system, such disturbances can be attenuated. This paper presents a workforce allocation system developed from the concepts of holonic manufacturing systems and PROSA reference architecture. The system is tested using a specifically developed C++ object-oriented software. Production data were gathered from the case study of a large computer manufacturing factory and used to simulate the proposed system. The paper then presents a comparison of the factory conventional method of workforce allocation and the proposed holonic workforce allocation system.

A ring-based interaction protocol for job allocation in manufacturing networks

This paper uses the holonic concept to develop a new model of job allocation in manufacturing networks. First, the conventional job allocation protocol of a manufacturing network was observed from an industrial case study and was modelled. Conventional practice uses a centralized approach to job allocation where a central company manages the job allocation process and the manufacturers of the network are, at that instant, passively receiving sub-orders. The research tries to introduce a ring-based interaction protocol, as an alternative model, which uses a distributed system paradigm. The interaction protocol consists of communication, along a virtual ring of manufacturers, which is uni-directional coupled with a peer-to-peer feedback mechanism and a supervisory body for conflict resolution between two manufacturers competing for the same sub-order. In this protocol, the candidates are virtually arranged into a ring according to rules and criteria. The candidates exhibit a ‘greedy’ behaviour whereby if they provide a feasible solution, the candidates take the maximum of the sub-order and the rest of the order list is circulated along the ring. If a candidate provides a sub-optimal solution, the next candidate can bid a better solution and inform the previous candidate that it has been outbidded. This is where the two candidates can compete to win the sub-order. The results herald the proposed ring-based model of job allocation as a powerful tool for improving lead time, cost and other pre-defined goals of a manufacturing network.

Agent Cooperation Mechanism for Decentralized Manufacturing Scheduling

This paper presents an agent cooperation mechanism for scheduling operations in a manufacturing network, while allowing manufacturers to absolutely control their scheduling activities. The study includes a thorough review of recent publications, a real-life industrial use case of a manufacturing network, an agent-based model of the network simulated with recursive porous agent simulation toolkit, the Muth and Thompson (MT10) scheduling data set, and the visualization of results in Microsoft Project. Results of a study of a four-layer cooperation mechanism showed that for the MT10 problem, manufacturer arrangement 0–5–7–2–3–8–1–9–6–4–0 was found to maximize the utilitarian social welfare of the manufacturing network. In terms of make-span, the network achieved a maximum of 1125 which was beyond the known optimal 930. Results suggest that manufacturers could express their scheduling goals and their preferences with whom they wanted to cooperate. These were measured by the time incentive and compatibility indicators. The latter could also be used to track the optimality loss in make-span optimization when implementing the decentralized scheduling approach in the context of manufacturing networks.

An ontology-based intelligent data query system in manufacturing networks

Abstract This paper investigates the development of an intelligent data query framework through the use of semantic web technologies for manufacturing purposes. The primary objectives of the ontology-based data query were to develop an efficient and scalable data interoperability and retrieval system; in order to find the most relevant query results with minimum message cost, most hits per query and least response time. This document explains the idea of ontology and the application of the same in the manufacturing domain. A computer simulation software was developed based on a real case study of distributed networks of manufacturing workshops. In this research, a semantic query algorithm was developed where query results are returned by investigating the semantic richness of each workshop. Results were compared with a semantic-free search mechanism based on key performance indicators. The results show the validity of the proposed model for efficient data query when compared to random search.