Daniël Telgen

Decentralized Autonomous-Agent-Based Infrastructure for Agile Multiparallel Manufacturing

This paper describes an agent-based software infrastructure for agile industrial production. This production is done on special devices called equip lets. A grid of these equip lets connected by a fast network is capable of producing a variety of different products in parallel. The multi-agent-based underlying systems uses two kinds of agents: an agent representing the product and an agent representing the equip let.

Production Scheduling in an Agile Agent-Based Production Grid

To meet the requirements of modern production, where short time to market, production driven by customer requirements and low cost small quantity production are important issues, we have been developing an agent-based software infrastructure for agile industrial production. This production is done on special devices called equip lets. A grid of these equip lets connected by a fast network is capable of producing a variety of different products in parallel. The multi-agent-based software infrastructure is responsible for the agile manufacturing. An important aspect of this software is the scheduling of the production. This paper describes a multi-agent-based solution for this problem. In our production system requests for products arrive at random times and every product must be completed before its deadline.

Embedded autonomous agents in products supporting repair and recycling

This paper describes a concept where products are equipped with agents that will assist in recycling and repairing the product. These so-called product agents represent the product in cyberspace and are capable to negotiate with other products in case of recycling or repair. Some product agents of broken products will offer spare parts, other agents will look for spare parts to repair a broken product. On the average this will enlarge the lifetime of a product and in some cases prevent wasting resources. Apart from reuse of spare parts these agents will also help to locate rare elements in a device, so these elements can be recycled more easily.

Automatic Structured Decomposition of Manufacturing Actions in an Agent-Based Manufacturing System

This paper describes how products definitions can be automatically decomposed, to be able to utilize a group of Reconfigurable Manufacturing Systems that offer generic and standardized services. This approach is performed by an agent-based architecture where all systems, both products and manufacturing systems, have an agent that represents them. This is seen as a new manufacturing paradigm, called Grid Manufacturing. Grids can ideally be set up in such a way that a range of products can be produced dynamically and in parallel, depending on the demand the logistic systems might even be configured in such a way that products can move freely between manufacturing systems to offer more flexibility. This paper describes the necessary steps for the decomposition of the product definition, the manufacturing process and the involved architecture. The chosen platforms to perform these tasks are Multi Agent Systems and Robot Operating System, to create a hybrid architecture that fully utilizes performance, stability and the dynamic behavior of these distributed platforms. As a result the manufacturing systems can easily be adapted to the current product demand, which shortens the time to market for new products and makes it easier to scale the production means. This is especially interesting for high mix, low volume production, which is becoming more important in the high-tech and creative industry.

Structured Analysis of Reconfigurable Manufacturing Systems

The realization of a short product-time-to-market is a key-challenge in the design of modern manufacturing equipment. Compression of lead-times for product design and manufacturing require a concurrent way of engineering. This implies that structural decisions about manufacturing-equipment need to be made when products are still under development. This introduces development risks; changes in the layout of production systems, due to late modifications in the product design, are inefficient for lead-time and cost. It is preferable that the production system can be designed in a ‘first-time-right’ fashion. Therefore, the architectural freeze of a manufacturing system is preferably pushed backwards in time to sustain modifications of the product design as long as possible. Reconfigurable Manufacturing Systems (RMS) have been developed for this purpose. With their modular structure, they can be integrated in a short period of time. Though this leaves more time for product development, it does not exclude the industrialization risks. Since configuration of equipment only works reliably if its process technology is well understood, it is needed that poorly functioning manufacturing processes are detected and addressed in an early stage. Only then, sufficient time is available for corrective actions to be taken. This paper presents a scientific framework to model the development of RMS. The method has the capability to uncover manufacturing risks during early development. In combination with RMS, the freeze of system architecture can indeed be pushed backwards in time. The method uses the ‘Structured Analysis Design Technique’ (SADT). The process risks, as outcome of the analysis process, are ranked using a Failure Mode Effect Analysis (FMEA) to determine the severity of their impact. It helps focussing on primary issues to be addressed. The method was applied to a true case; the development of a RMS for cell Phone lenses. The industrialization process may be considered successful. By application of this approach, engineers profit of a complete overview of what actions need to be taken and the effects if these actions are omitted. The method can also be used to inform higher management, to increase understanding of the cause and effect of management decisions related to manufacturing.

Agile product manufacturing by dynamically generating control instructions

Grid Manufacturing (GM) is a new production paradigm, based upon the use of standardized and modular Reconfigurable Manufacturing Systems (RMS). In GM all systems have a virtual counterpart that actsautonomously, this includes both complete manufacturing systems and the products. The control system required for this approach is based upon a distributed and hybrid architecture, using agent technology. An important aspect in the paradigm is the product manufacturing description. This paper introduces the concept of an architecture where the control of the manufacturing is abstracted from the product manufacturing blueprint. A product is delineated step by step by specific services in the grid. The proposed system increases flexibility twofold, first by enabling abstraction of products parts and second by dynamically using manufacturing means.

A Versatile Agile Agent-Based Infrastructure for Hybrid Production Environments

Abstract To meet the requirements of modern production, where short time to market, production driven by customer requirements and low cost small quantity production are important issues, we have developed an agent-based software infrastructure for agile industrial production. This production is done on special devices called equiplets. A grid of these equiplets connected by a fast network is capable of producing a variety of different products in parallel. The multi-agent-based software infrastructure is responsible for the agile manufacturing. This agent-based model also fits well in a production environment where the production is done by humans instead of equiplets. This is what we call a hybrid production environment.

A Multiagent-Based Agile Work Distribution System

This paper describes an agent-based and web-based system to instruct workers to build a product or to perform certain actions. The making of a product or actions to be performed is conducted by a so-called product agent that knows what should be done. In many cases the work to be done consists of a sequence or set of steps. The workers advertise their capabilities or steps they can perform and a product agent selects workers based on the steps the workers have advertised and distributes the work among the workers. Both workers and products are represented by agents that interact in a multiagent system. A simple example implementation is also presented as a proof of concept.

A Versatile Simulation Model of Agile Production by Autonomous Systems

Complex situations and systems can be studied by using adequate models in simulation. An important aspect of models and the simulation software is the ability to use a wide range of possible input parameters. The simulation described in this paper is based on agile manufacturing by using transport robots and cheap reconfigurable production platforms, called equiplets. This setup makes agile manufacturing of different products in parallel possible.The simulation showed the maximum load of such a production environment as well as a proof of concept for the distributed approach for transport used.

Implementation of an Agent-Based Manufacturing Grid

This paper describes the implementation of an agile autonomous agent-based manufacturing system based on a grid. This grid contains production machines, represented by agents, capable to perform certain production steps. Products to be made are also represented by agents. Many different products can be made in parallel, each product having its own sequence of production steps. The whole manufacturing is based on interaction of agents living in a distributed environment. This paper explains the basic design considerations and includes a simple example as a proof of concept.