Jürgen Bode

Neural networks for cost estimation: Simulations and pilot application

Neural networks in a multilayer perceptron architecture are able to classify data and approximate functions based on a set of sample data (curve fitting). These properties are used to investigate experimentally the applicability of neural networks for cost estimation in early phases of product design. Experiments are based on pilot cost data from a manufacturing company. In addition, artificially created simulative data are used for benchmarking. The cost estimation performance is compared to conventional methods, i.e. linear and non-linear parametric regression. Neural networks achieve lower deviations in their cost estimations. Beyond the use of standard neural architectures, simple modifications for a performance improvement are suggested and tested. Finally, a profile for situations where neural networks are appropriate is derived from the results.

Cost engineering with quality function deployment

Conventional Quality Function Deployment (QFD) is technically one-sided. Prioritization of technical attributes, if carried out at all, attempts to maximize customer satisfaction without considering the costs incurred. However, product design is usually a techno-economic process, hence there is always a tradeoff between quality goals and limited budgets. Based on a prioritization method suggested by Wasserman [1], this paper integrates design costs into the QFD framework. This proposed approach enables designers to optimize product development resources towards customer satisfaction and conduct analytical investigations to facilitate decision making in product design and development.

Application of multiagent systems in project management

Abstract Global markets often require the coordination of firms across the world to gain maximum competence. The organization structure and the management of a firm tend to be project-oriented, and members of a project team are distributed throughout a network. In this paper, we use multi-agent systems as a technique to support project management in a distributed environment. Activity agents and resource agents represent activities and resources in a project. Service agents are automated experts for special project management tasks. This paper presents methods to schedule activities and resolve resource conflict by message exchanging and negotiation among agents. Finally, a prototype of project management tool is implemented, which uses a network of agents to provide services to team members distributed throughout Internet.

A framework of decision‐support systems in advanced manufacturing enterprises ‐ a systems view

Analyses the environment and characteristics of an advanced manufacturing system (AMS). It is an open system with a multi‐layer structure and a self‐organizing ability capable of responding to a continuous changing and unpredictable environment in this information age. Based on the analysis, summarizes the requirements of decision processes in a typical AMS, and presents a framework of a decision‐support system (DSS) in an advanced manufacturing enterprise. Outlines the conceptual modelling of the system, explains the work carried out by an inter‐disciplinary team composed of researchers from the 863/CIMS/I‐MADIS, a national hi‐tech R&D programme in China and a joint research programme in computer integrated manufacturing management between the City University of Hong Kong and Tsinghua University, Beijing. 863/CIMS is one of the subject themes under the auspices of automation technology of the National High Technology Research and Development Programme of China launched by the government in March 1986.

Multi-agent system for cost estimation

Abstract The cost estimation now is a vital concern of every manufacturing firm, and many methods are available for cost estimation. In this paper , we build an novel architecture of multiagent system with each agent representing one kind of cost estimation method (distributed multi-agent system). The system first classifies the users requests , then makes the expert agents to cooperate and compete to fulfill the task. The model is based on a fuzzy classification of cost estimation methods and a dynamic optimization of structure to make the agents to learn from the past (based on agent reputation). In this way, the amount of information exchange is considerably reduced.

A multiagent system for the support of concurrent engineering

The concurrent design of a product involves the cooperation of many functions such as manufacturing, logistics, marketing, and sales. It can be represented as a multiagent system with each agent representing a kind of expertise. Each agent has its own evaluation criteria and egoistic goals. Decision-making leads to problems typical for concurrent engineering, in which all related processes are simultaneously taken into consideration. In this paper, we discuss some characteristics of conflict, conflict type identification, as well as conflict resolution. A system architecture and an agent structure are proposed. We also propose a novel scheme named distributed dynamic conceptual tree to overcome the difficulty of understanding between agents during the problem solving process and reduce the amount of communications between agents. An demonstration example validating our work is presented.

Neural networks in new product development

Presents the results of the application of back propagation three layer perceptron to cost estimation problems in design. Typical requirements of neural networks to be used in new product development are derived. Network construction has to consider the small number of training sets as a major characteristic of design problems.

Cost estimation in new product development: Knowledge pluralism, knowledge granularity, and agent based representation

Presents a distributed model based on the contract net approach to support cost estimation during all phases of product design. Several methods are available for cost estimation (knowledge pluralism), each one represented by one agent. Agents compete and cooperate to solve costing problems. Unlike common multi agent architectures agent construction follows data accuracy and data availability (knowledge granularity) of the method to be applied.

CODESS-a system for decision support in the conceptual phase of new product development: concepts and application

The goal of the customer-oriented design support system (CODESS) is to implement quality function deployment (QFD) into other information systems more effectively and provide decision support in the conceptual phase of new product development. This paper describes the CODESS approach to design decision support and discusses a data model which facilitates development and maintenance of CODESS and plays a key role in the integration of CODESS with the corporate information system. In the final section, we report on the results of implementing and applying the system in a pilot company, and derive general findings on effective and efficient support of design decisions in the conceptual phase of new product development.

Ethisches Verhalten in Betrieben – leicht gefordert, schwer erreicht? Was wir aus dem Beispiel des Fußballs lernen können

Am 18. November 2009 spielten die Fusballnationalmannschaften Frankreichs und Irlands im Stade de France gegeneinander. Das Hinspiel war fur Frankreich 1:0 ausgegangen, nun stand im Ruckspiel die Entscheidung an, welches Team sich fur die Weltmeisterschaft 2010 in Sudafrika qualifizieren wurde.