Use Cases of Discrete Event Simulation

Variance reduction techniques have been shown by others in the past to be a useful tool to reduce variance in Simulation studies. However, their application and success in the past has been mainly domain specific, with relatively little guidelines as to their general applicability, in particular for novices in this area. To facilitate their use, this study aims to investigate the robustness of individual techniques across a set of scenarios from different domains. Experimental results show that Control Variates is the only technique which achieves a reduction in variance across all domains. Furthermore, applied individually, Antithetic Variates and Control Variates perform particularly well in the Cross-docking scenarios, which was previously unknown.

The planning of earthworks represents a complex task. The use of different machine configurations as well as alternative scenarios in the site layout (e.g. transport routes and temporal storage areas) must be evaluated and dimensioned consistently. Wrong decisions can lead to delays or an uneconomic solution and hence increase the costs and project duration. In practice, this planning process is based on the experience and knowledge of the persons in charge; however, decision support tools are not used in the planning of excavation and transportation equipment despite their central importance. Therefore an approach has been developed to support the planning of construction processes in earthworks by applying discrete event simulation. For this purpose, methods for calculating the performance of earthmoving equipment were extended based on statistical components, adapted for simulation, and implemented in a module library. Furthermore, the simulation tool has been coupled with a mathematical optimization procedure to reduce the cost of transport in earthworks by minimizing haul times.

Simulation analyses subdivide themselves conveniently into two major categories: discrete-event simulation and continuous simulation (Zeigler, Praehofer, and Kim 2000). Continuous simulation studies processes amenable to analysis using differential and difference equations, such as stability of ecological systems, chemical synthesis, oil refining, and aerodynamic design. Discrete-event simulation studies processes in which many of the most important variables are integer values, and hence not amenable to examination by continuous equations. Such processes almost invariably involve queuing, and the variables of high interest include current and maximum queue lengths, number of items in inventory, and number of items processed by the system. Many of the integer values are binary; for example, a machine is in working order or down, a worker is present or absent, a freight elevator is occupied or vacant. Processes with these characteristics are common in manufacturing, warehousing, transport, health care, retailing, and service industries.

Energy and resource efficiency emerge as strategic objectives in the operation of discrete manufacturing systems. In the future, energy consumption will have to be evaluated early during the planning phases, requiring the application of simulation technology. This has, until now, not been implemented into the tools of the digital factory supporting this phase of the product lifecycle.

MAGNA STEYR aims to establish digital planning in all important areas. In the field of BIW (Body In White) the digital process planning is already a reality. Starting from a digital product model, welding process are planned completely in digital form. Process simulation and offline robot programming safeguard the planning. With the connection of the digital process planning and discrete event simulation MAGNA STEYR took an important step towards realizing the digital factory.

Manufacturing industries across the globe face numerous challenges to become 100% efficient but each and every industry has its own constraints / problems with their functional system to achieve 100% excellence. End of the month syndrome is one of the major problems almost all manufacturing industries face with the ever growing demand and the competition around. Manufacturers find it really difficult to achieve their potential if they produce more than 25% of their monthly shipment plan in the last week of the month or more than 33% of their quarterly shipment plan in the last month of the quarter. Companies that live with the "end-of-the-month-crunch will be burdened with premium freight, internal expediting, overtime costs, and production inefficiencies that will crush their bottom line goals. But effective upfront planning and timely execution can make the “end-of-the-month-crunch" a bad memory and eliminate those profit killers. The causes for end of the month syndrome are raw material constraints and production inefficiencies, last minute product changes, stoppage and machine down time in manufacturing line etc. Manufacturing industries can analyze these challenges through the application of modeling and simulation technique with the existing system and try out various “what-if scenarios” (sensitivity analysis) without any physical changes to the existing process & thus find a solution to all those problems leading to End of the Month Syndrome.

The increasing mastery of the instrument Discrete Event Simulation and increasing detailing of the simulation models open up new fields for the simulation. The following article deals with the use of discrete event simulation in the field of commissioning of production lines. This type of modeling requires the inclusion of sensors and actuators of the manufacturing facility. Our experience shows that it is well worth the effort. Essential coordination with the development of automation can be integrated in the planning process. The simulation helps to find a common language with all people involved in the development.

This paper explores the use of simulation for the optimization of highly flexible production plants. Basis for this work is a model of a real shoe production plant that produces up to 13 different styles concurrently, resulting in maximum 11 different production sequences. The flexibility of the plant is ensured by organizing the process in a sequence of so-called work islands, using trolleys to move shoes between them. Depending on production needs one third of the operators are reallocated. The model considers the full complexity of allocation rules, assembly flows and production mix. Analyses were performed by running use cases, from very simple (providing an insight in basic dynamics) up to complex (supporting the identification of interaction effects and validation against reality). Analysis gave insight in bottlenecks and dependencies between parameters. Experiences gained distilled in guidelines on how simulation can support the improvement of highly flexibly organized production plants.

High variability is a harmful factor for manufacturing performance that may be originated from multiple sources and whose effect might appear in different temporary levels. The case study analysed in this chapter constitutes a paradigmatic case of a process whose variability cannot be efficiently controlled and reduced. It also displays a complex behaviour in the generation of intermediate buffers. Simulation is employed as a tool for detailed modelling of elements and variability components capable of reproducing the system behaviour. A multilevel modelling approach to variability is validated and compared to a conventional static model in which process parameters are kept constant and only process cycle dependant variations are introduced. Results show the errors incurred by the simpler static approach and the necessity of incorporating a time series model capable of simulating the autocorrelation structure present in data. A new layout is proposed and analysed by means of the simulation model in order to assess its robustness to the present variability. The new layout removes unnecessary process steps and provides a smoother response to changes in the process parameters.

Modeling and Simulation techniques are powerful tools for evaluating best layout option by analyzing key performance indicators of a given process. A simulation technique for layout validation has its own unique benefit because the element of risk associated is almost zero. By sensitivity analysis potential process improvement strategies can be identified, evaluated, compared and chosen in a virtual environment much before the actual implementation and this helps in better decision making.

Initially relevant data required was collected verified and cleaned using various statistical tools. After building the initial model an “AS-IS” model evolved, as the results were presented and discussed highlighting the pit falls in the current layout which affects the performance of the plant with process owners. At the analysis stage various “WHAT-IF” scenarios were identified and evaluated so as to identify the best alternative depending upon the performance measures which have the most significant improvement.

This would, hence, become a prerequisite for management in arriving at a better decision after evaluation of various alternative results obtained from the simulation.

Designing large-scale High-Performance Computing (HPC) systems, including architecture design space exploration and performance prediction, is a daunting task that can benefit enormously from discrete event simulation techniques, as the interactions between the various components of such a system generally render analytic approaches intractable. The work described in this chapter specifically deals with end-to-end, full-system simulation, as opposed to simulation of individual components or nodes. The tools described here can be used in the design phase of a new HPC system to optimize system design for a given set of workloads, or to create performance forecasts for new workloads on existing systems.

We have taken a network-centric approach, as the scale of current high-end HPC systems is in the range of hundreds of thousands of processing cores, so that the impact of the communication among so many cores will be a key factor in determining overall system performance. To this end, we developed an Omnest-based simulation environment that enables studying the impact of an HPC machine’s communication subsystem on the overall system’s performance for specific workloads.

Full system simulation at an abstraction level that still maintains a reasonably high level of detail is infeasible without resorting to parallel simulation, the main limiting factors being simulation run time and memory footprint. By applying Parallel Discrete Event Simulation techniques, the power of modern parallel computers can be exploited to great effect to perform these kinds of simulations at large scales.

This chapter deals with the modular library ‘Automotive’ (in original VDA Automotive Bausteinkasten) of the software Plant Simulation with the focus on point-oriented elements from this library. First, a general introduction to specific modular libraries in Plant Simulation, their purpose, way of use and limits is presented. A brief description of the library ‘Automotive’, its historical as well as current development, structure and field of use follows. The core of this chapter presents two sample models which show the use of the library ‘Automotive’. The aim is to give the reader insight into the variety of the modules and objects of the library ‘Automotive’ which enable the user to efficiently simulate various processes we can encounter in the automotive industry.

In this chapter, we illustrate the use of discrete event simulation to evaluate how dynamic planning methodologies can be best applied for the collection of waste from underground containers. We present a case study that took place at the waste collection company Twente Milieu, located in The Netherlands. Even though the underground containers are already equipped with motion sensors, the planning of container emptying’s is still based on static cyclic schedules. It is expected that the use of a dynamic planning methodology, that employs sensor information, will result in a more efficient collection process with respect to customer satisfaction, profits, and CO₂ emissions. In this research we use simulation to (i) evaluate the current planning methodology, (ii) evaluate various dynamic planning possibilities, (iii) quantify the benefits of switching to a dynamic collection process, and (iv) quantify the benefits of investing in fill-level sensors. After simulating all scenarios, we conclude that major improvements can be achieved, both with respect to logistical costs as well as customer satisfaction.

Production processes in the chemical industry are to a large extend not discrete but continuous. Hence, the application of discrete-event simulation (DES) in this field is not as widespread as in discrete manufacturing. In order to apply DES methodology to chemical production processes, continuous aspects have to be covered sufficiently. This contribution briefly introduces and discusses combined discrete-continuous simulation approaches and illustrates the potential of the methodology using three cases of a leading German chemical company from supply chain optimization to the shop floor.

In the brewing industry the quantities to be produced for each product are specified in weekly meetings of planning. Afterwards, manually detailed planning and resource scheduling is carried out by a production specialist, who usually takes basic applications developed on MS Excel as a planning support.

A disadvantage of this planning process is the high spending time because of the high complexity of hundreds of constraints involved in the production process and the need of having expertise to understand how the process might change along the week because of many related biological processes.

Through application of simulation with Plant Simulation for production planning and resource scheduling the stated disadvantages can be avoided. This simulation is oriented to be a planning tool that automatically generates the production schedule on the basis of current stocks, master production schedule and minimum lot sizes, which are included taking into account all manufacturing restrictions.

The user can configure plant parameters, stock levels and week production orders. The scheduling tool thus generates the production schedule in a really short time and makes possible evaluating several production scenarios and makes the best decision to optimize plant utilization, stock levels and throughput times.

In this chapter is presented a scheduling tool for breweries based on a simulation of a real plant, the development and the benefits achieved using it in the real process of planning.

The book chapter presents two case studies that consequently use the computer-aided simulation in combination with the optimization. The optimization follows the search of the best solution for a given optimization problem. Case study 1 introduces a special procedure for the determination of the number of machines in production systems. Thereby the optimization is combined with a cost simulation. It shows that, with this procedure, very good solutions can be found automatically and concerning a specific problem. Case study 2 deals with the order controlling in Car Assembly with the Aid of Optimizers. The modeling had to consider that a lot of flexible parameters were needed to ensure enough planning roam. A main goal was to determine the computational achievable “right” production sequence. The hand-made production program should be optimized by the simulation. Both case studies present possibilities and potentials of the computer-aided simulation combined with the optimization.