Abstracts of the seminar papers

Gautam Mitra (Director Center for the Analysis of Risk and Optimisation Modelling Applications, Professor of Computational Optimisation and Modeling, Brunel University, Mathematical Sciences, Cleveland Road, Uxbridge, Middlesex UB8 3PH, England)
E-mail: gautam.mitra@brunel.ac.uk

Emerging architecture of tools and components for quantitative modelling and decision support.

Quantitative models for industrial and management applications continue to co-evolve with developments in software systems, business needs and management thinking.  We examine the role of management models in this context.  We also consider recent technology developments and analyse how these address the needs of the end user community. The interaction of analytic data modelling (datamarts) with decision modelling (optimisation) and descriptive modelling (simulation) are considered. Finally, we propose a modelling architecture which may bring together a wide class of established models. These concepts will be illustrated using deterministic as well as stochastic optimisation models and simulation models for results analysis. Web enabled applications in the domains of finance and supply chain logistics will be discussed.
 
 

Jan Bisschop (Paragon Decision Technology B.V., P.O. Box 3277, 2001 DG  Haarlem)
E-mail: jan.bisschop@paragon.nl
Ruud Brekelmans (CentER Applied Research, P.O. Box 90153, 5000 LE  Tilburg)
E-mail: r.c.m.brekelmans@uvt.nl

The AIMMS multi-agent technology and a management chain game

During the first part of the presentation there will be an introduction to the Advanced Integrated Multi-dimensional Modeling Software system (AIMMS) and to the AIMMS Multi-Agent technology for distributed decision making in particular.  This introduction will focus on basic agent concepts together with some illustrative examples.  The introduction will also illustrate the way agents have been implemented in AIMMS.  During the second part of the presentation the NetChain game, a KLICT/ICES project, is introduced as an application of the AIMMS Multi-Agent technology.  In this supply chain management game the participants have to make upstream and downstream contracts for the product-market combinations in which they want to be active. This is carried out by announcing offers and holding contract negotiations.  In addition, production and shipments take place in real-time, and the participants have to react to changes in the market situation and ensure that contracts are being satisfied.  Compared to other games, the innovative part lies in the combined turn based (strategic negotiations) and real-time decisions (day-to-day operations). The AIMMS Multi-Agent technology has contributed to the successful implementation of this game.
 
 

Rienk Bijlsma (Simulation & Scheduling Expert at Rockwell Automation, Rivium 1e Straat 23, 2909 LE Capelle a/d IJssel)
E-mail: rbijlsma@software.rockwell.com

The future in improving design and operations with simulation technology

Simulation Technology has been used for decades to support decision-making in design of new facilities and systems, as well as improving existing ones. Its use is however fragmented and very often not embedded in the organization. Organizations that do use this technology extensively, value the benefits properly, but still are unable to embed simulation models in their operational systems. Recent software developments can and may result in a shift of the proper placement of this technology within the organization. The following developments can lead to this shift and will be handled during this presentation:
- Data availability and connectivity to Legacy Systems
- User Friendliness and Communication
- Execution Speed and Robustness
- Real Time Execution and Emulation
- Simulation of Control Code (PLC)
- Joined Model Building, version control
 
 

Victor Allis (Director Quintiq, P.O. Box 264, 5201 AG ‘s Hertogenbosch)
E-mail: victor.allis@quintiq.com

The Holy Grail of advanced planning and scheduling systems

The demand for APS systems has been fueled by the awareness that in many companies the “daily puzzle” has grown more complex over the years. The daily puzzle that must be solved determines the execution of operations of the company. For a production company this involves determining who will produce which (production) orders on which machines, in which order/batches/runs, using which materials at what time. For a logistical operation this involves determining who will transport what products/people from where to where in what combination, at what time.
The complex daily puzzle is typically extremely important for a company: finding a better solution to it, often has a direct and significant influence on the bottom line. For example, a production company that solves its daily puzzle in a better way may see significant positive effects on key performance indicators (KPIs) such as its production output, its delivery performance, its material yield etc. Also, a logistical company may see significant positive effects on its punctuality, its average resource (truck/plane/crew) utilization, fuel costs, etc.
Traditional transaction systems (e.g. ERP systems, TMS systems) typically provide no solution for these complex puzzles. Instead, they typically plan at a rather abstract level, schedule against infinite capacity, or even if they schedule against finite capacity, fail to take many of the relevant details into account, resulting in a sub-optimal schedule or planning. This has sparked the development of what has become known as APS systems.
The question now is, what the key element is of a successful APS implementation. Customers often believe that it must come from a fully automated scheduler. Sometimes they believe that automation is the Holy Grail, and that APS stands for Automated Planning and Scheduling. This customer belief may be reinforced by the computer scientist or applied mathematician, for whom the most challenging task is to program an algorithm to automatically solve a scheduling puzzle.
We believe that, although automated schedulers using clever algorithms are important, they are not the main key to success. The heart of the solution must be found in creating the right model for each specific challenge. The Quintiq vision is based on this belief and will be presented in the lecture.
 
 

Fred Jansma (Manager Software Development, Incontrol Enterprise Dynamics, Planetenbaan 21, 3606 AK Maarssen)
E-mail: fred.jansma@incontrol.nl

Analyzing and controlling your business with enterprise dynamics

Business control is heavily relying on software systems today. Most of our information has been processed by a computer somewhere. Often we do not know by which computer or software and who designs the information systems. Still many people take decisions based on the information they have gathered without questioning.
Maybe not realizing that business control has shifted to a great extend into the area of software control. Software systems have evolved from just a tool into our main source of information. And software systems control many operational processes.
Software quality, capability and reliability have become key factors to every researcher and business manger today.
Operations research has their own set of tools. But more and more the OR tools and business control systems are becoming the same.
Enterprise Dynamics has integrated modeling, simulation, visualization and control into one -open- environment. The same Enterprise Dynamics software is used in OR and business operations.
Based on a real life case study at the port of Amsterdam we will demonstrate how OR and business control is implemented with Enterprise Dynamics highlighting the architecture and features of our latest software.