| 1. |
- Lidberg, Simon, MSc. 1986-, et al.
(author)
-
A Knowledge Extraction Platform for Reproducible Decision-Support from Multi-Objective Optimization Data
- 2022
-
In: SPS2022. - Amsterdam; Berlin; Washington, DC : IOS Press. - 9781643682686 - 9781643682693 ; 21, s. 725-736
-
Conference paper (peer-reviewed)abstract
- Simulation and optimization enables companies to take decision based on data, and allows prescriptive analysis of current and future production scenarios, creating a competitive edge. However, it can be difficult to visualize and extract knowledge from the large amounts of data generated by a many-objective optimization genetic algorithm, especially with conflicting objectives. Existing tools offer capabilities for extracting knowledge in the form of clusters, rules, and connections. Although powerful, most existing software is proprietary and is therefore difficult to obtain, modify, and deploy, as well as for facilitating a reproducible workflow. We propose an open-source web-based application using commonly available packages in the R programming language to extract knowledge from data generated from simulation-based optimization. This application is then verified by replicating the experimental methodology of a peer-reviewed paper on knowledge extraction. Finally, further work is also discussed, focusing on method improvements and reproducible results.
|
|
| 2. |
- Lidberg, Simon, MSc. 1986-
(author)
-
Evaluating Fast and Efficient Modeling Methods for Simulation-Based Optimization
- 2021
-
Licentiate thesis (other academic/artistic)abstract
- As the industry in general, and the automotive industry in particular, is transforming -- due to new technologies and changes in market demands through electrification, digitalization, and globalization -- maintaining a competitive edge will require better predictions. Better predictions of production performance allows companies to capitalize on opportunities, avoid costly mistakes, and be proactive about change.A commonly used tool in manufacturing for the prediction of production performance is discrete-event simulation. In combination with artificial intelligence methods such as multi-objective optimization, in literature often referred to as simulation-based optimization, and knowledge extraction, bottlenecks in the production process can be identified and recipes for optimal improvement order can be obtained. These recipes support the decision-maker in both understanding the production system and improving it optimally in terms of resource efficiency and investment cost. Even though the use of simulation-based optimization is widespread on the production line level, use on the factory level is more scarce. Improvements on the production line level, without a holistic view of factory performance, can be suboptimal and may only lead to increased storage levels instead of increased output to the customer.The main obstacle for applying simulation-based optimization to the factory level is the complexity of its constituent parts, i.e., detailed production line models. Connecting several detailed production line models to create a factory model results in an overly complicated, albeit, accurate model. A single factory model running for one minute would equate to almost 140 days required for an optimization project, too long to provide decision-support relevant to manufacturing decision-making. This can be mitigated by parallel computing, but a more effective approach is to simplify the production line models to decrease the runtime while trying to maintain accuracy. Model simplification methods are approaches to reduce model complexity in new and existing simulation models. Previous research has provided an accurate and runtime efficient simplification method by use of a generic model structure built by common modeling components. Although the method seems promising in a few publications, it was lacking external and internal validity.This project presents simulation-based optimization on the factory level enabled by a model simplification method. By following the design science research methodology, several case-studies mainly in the automotive industry identify issues with the current implementation, propose additions to the method, and validates them.
|
|
| 3. |
- Lidberg, Simon, MSc. 1986-, et al.
(author)
-
Model Simplification Methods for Coded Discrete-Event Simulation Models : A Systematic Review and Experimental Study
-
Other publication (other academic/artistic)abstract
- In an increasingly competitive market due to customer demands of customization and an increasing rate of new product variant introductions, companies need to explore new tools to support them to better predict and optimally re-configure their production networks. In terms of the factory flow level, discrete-event simulation and simulation-based optimization represent such kinds of tools available at the disposal of production engineers or managers. For a complex factory consisting of multiple production lines, creating detailed simulation models of these lines and connecting them together can be used for optimization, but the computational complexity can be prohibitively large for acquiring results in time. Model simplification methods can be utilized to reduce the computational complexity of a model. In this study, a systematic literature review is conducted with the aim of identifying simplification methods for coded models, characteristics of the detailed model, type of industry, motivation, and validation measures. Based on the results of the literature review an experimental study wherein the limits of a specific simplification method are analyzed. We compare the output of a dynamically created model with the output of a simplified representation. A correlation can be observed between outputs for medium to large lines, but for smaller lines, there is a larger discrepancy. The simplification method allows for the reduction in simulation runtime, enabling simulation-based optimization of large lines or interconnected simplified models forming a production network, i.e., a factory, to be optimized and analyzed more efficiently, leading to competitive advantages for companies.
|
|
| 4. |
- Morshedzadeh, Iman, 1979-, et al.
(author)
-
Multi-level management of discrete event simulation models in a product lifecycle management framework
- 2018
-
In: Procedia Manufacturing. - : Elsevier. - 2351-9789. ; 25, s. 74-81
-
Journal article (peer-reviewed)abstract
- Discrete event simulation (DES) models imitates the behavior of a production system. Models can be developed to reflect different levels of the production system, e.g supply chain level or manufacturing line level. Product Lifecycle Management (PLM) systems have been developed in order to manage product and manufacturing related data. DES models is one kind of product lifecycle’s data which can be managed by a PLM system. This paper presents a method and its implementation for management of interacting multi-level models utilizing a PLM system.
|
|
| 5. |
- Frantzén, Marcus, 1981-
(author)
-
A real-time simulation-based optimisation environment for industrial scheduling
- 2013
-
Doctoral thesis (other academic/artistic)abstract
- In order to cope with the challenges in industry today, such as changes in product diversity and production volume, manufacturing companies are forced to react more flexibly and swiftly. Furthermore, in order for them to survive in an ever-changing market, they also need to be highly competitive by achieving near optimal efficiency in their operations. Production scheduling is vital to the success of manufacturing systems in industry today, because the near optimal allocation of resources is essential in remaining highly competitive. The overall aim of this study is the advancement of research in manufacturing scheduling through the exploration of more effective approaches to address complex, real-world manufacturing flow shop problems. The methodology used in the thesis is in essence a combination of systems engineering, algorithmic design and empirical experiments using real-world scenarios and data. Particularly, it proposes a new, web services-based, industrial scheduling system framework, called OPTIMISE Scheduling System (OSS), for solving real-world complex scheduling problems. OSS, as implemented on top of a generic web services-based simulation-based optimisation (SBO) platform called OPTIMISE, can support near optimal and real-time production scheduling in a distributed and parallel computing environment. Discrete-event simulation (DES) is used to represent and flexibly cope with complex scheduling problems without making unrealistic assumptions which are the major limitations of existing scheduling methods proposed in the literature. At the same time, the research has gone beyond existing studies of simulation-based scheduling applications, because the OSS has been implemented in a real-world industrial environment at an automotive manufacturer, so that qualitative evaluations and quantitative comparisons of scheduling methods and algorithms can be made with the same framework. Furthermore, in order to be able to adapt to and handle many different types of real-world scheduling problems, a new hybrid meta-heuristic scheduling algorithm that combines priority dispatching rules and genetic encoding is proposed. This combination is demonstrated to be able to handle a wider range of problems or a current scheduling problem that may change over time, due to the flexibility requirements in the real-world. The novel hybrid genetic representation has been demonstrated effective through the evaluation in the real-world scheduling problem using real-world data.
|
|
| 6. |
- Frantzén, Marcus, 1981-, et al.
(author)
-
Dynamic maintenance priority of a real-world machining line
- 2016
-
In: Proceedings of the 7th Swedish Production Symposium.
-
Conference paper (peer-reviewed)abstract
- To support the shop-floor operators, decision support systems (DSS) are becoming more and more vital to the success of manufacturing systems in industry today. In order to get a DSS able to adapt to disturbances in a production system, on-line data are needed to be able to make optimal or near-optimal decisions in real-time (soft real-time). This paper investigates one part of such a system, i.e. how different priorities of maintenance activities (planned and unplanned) affect the productivity of a production system. A discrete-event simulation model has been built for a real-world machining line in order to simulate the decisions made in subject to disturbances. This paper presents a way of prioritizing operators and machines based on multiple criteria such as competence, utilization, distance, bottleneck, and Work-In-Process. An experimental study based on the real-world production system has shown promising results and given insights of how to prioritize the operators in a good way. Another novelty introduced in this paper is the use of simulation-based optimization to generate composite dispatching rules in order to find good tradeoffs when taking a decision of which machine or operator to select.
|
|
| 7. |
- González Chávez, Clarissa Alejandra, 1994, et al.
(author)
-
Achieving Sustainable Manufacturing by Embedding Sustainability KPIs in Digital Twins
- 2022
-
In: Proceedings - Winter Simulation Conference. - 0891-7736. ; 2022-December, s. 1683-1694
-
Conference paper (peer-reviewed)abstract
- The manufacturing industry requires highly flexible and dynamic production lines that shift from conventional mass production to cover the requirements and fulfill demands. Customized production may reduce production waste but has not been studied to a wide extent. The advancement of digital technologies, e.g., Digital Twins, enable factories to collect real-time data. Also, they can enable remote monitoring of the production processes by establishing bi-directional flows of data between the physical and virtual spaces. This study draws its sight to the potential of digital manufacturing to improve sustainability in production systems by making use of Digital Twins. This research work performs a literature review and identifies suitable KPIs for a DES model and evaluates the impact in a drone factory in four scenarios that test final assembly processes. The findings of this work can pose a first step toward the future development of a digital twin.
|
|
| 8. |
- Lidberg, Simon, 1986-, et al.
(author)
-
Applying Aggregated Line Modeling Techniques to Optimize Real World Manufacturing Systems
- 2018
-
In: Procedia Manufacturing. - : Elsevier. - 2351-9789. ; 25, s. 89-96
-
Journal article (peer-reviewed)abstract
- The application of discrete event simulation methodology in the analysis of higher level manufacturing systems has been limited due to model complexity and the lack of aggregation techniques for manufacturing lines. Recent research has introduced new aggregation methods preparing for new approaches in the analysis of higher level manufacturing systems or networks. In this paper one of the new aggregated line modeling techniques is successfully applied on a real world manufacturing system, solving a real-world problem. The results demonstrate that the aggregation technique is adequate to be applied in plant wide models. Furthermore, in this particular case, there is a potential to reduce storage levels by over 25 %, through leveling the production flow, without compromising deliveries to customers.
|
|
| 9. |
- Ma, Andrew, et al.
(author)
-
Anarchic manufacturing : Distributed control for product transition
- 2020
-
In: Journal of manufacturing systems. - : Elsevier. - 0278-6125 .- 1878-6642. ; 56, s. 1-10
-
Journal article (peer-reviewed)abstract
- Manufacturers are poorly equipped to manage product transition scenarios, when moving from one product to another. Most tools consider a mature system, yet during transition and ramp up disturbances and inefficiency are common. The traditional methods, using centralised planning and control structures are too rigid and often resort to simple dispatch heuristics in this highly dynamic environment. Distributed systems have been proposed to leverage their self-organising and flexible traits to manage highly volatile and complex scenarios. Anarchic manufacturing, a free market based distributed planning and control system, delegates decision-making authority and autonomy to system elements at the lowest level; this system has previously been shown to manage job and flowshop style problems. The system has been adapted to use a dynamic batching mechanism, where jobs cooperate to benefit from economies of scale. The batch enables a direct economic viability assessment within the free market architecture, whether an individual machine should changeover production to another product type. This profitability assessment considers the overall system state and an agent's individual circumstance, which in turn reduces system myopia. Four experiments, including a real-world automotive case study, evaluate the anarchic manufacturing system against two centralised systems, using three different ramp-up curves. Although not always best performing against centralised systems, the anarchic manufacturing system is shown to manage transition scenarios effectively, displaying self-organising and flexible characteristics. The hierarchical system was shown to be impeded by its simplifying structure, as a result of structural rigidity. © 2020 The Society of Manufacturing Engineers
|
|
| 10. |
- Pehrsson, Leif, 1970-, et al.
(author)
-
Aggregated models for decision-support in manufacturing systems management
- 2021
-
In: International Journal of Manufacturing Research. - : InderScience Publishers. - 1750-0591 .- 1750-0605. ; 16:3, s. 217-240
-
Journal article (peer-reviewed)abstract
- Many industrial challenges can be related to the setup of manufacturing plants and supply chains. While there are techniques available for discrete event simulation of production lines, the opportunities of applying such techniques on higher manufacturing network levels are not explored to the same extent. With established methods for optimisation of manufacturing lines showing proven potential in conceptual analysis and development of production lines, the application of such optimisation methods on higher level manufacturing networks is a subject for further exploration. In this paper, an extended aggregation technique for discrete event simulation of higher level manufacturing systems is discussed, proposed, tested, and verified with real-world problem statements as a proof of concept. The contribution of the new technique is to enable the application of DES models, with reasonable computational requirements, at higher level manufacturing networks. The proposed technique can be used to generate valuable decision information supporting conceptual systems development.
|
|
