The application of real-time spatial optimization in a Spatial Decision Support System, utilizing Geographic Information Science and Technology, is shown using the example of Wood Supply Chain management. Interdisciplinarity is a key concept, because Geographic Information Science, Operations Research and Spatial Decision Support Systems are bundled together. Service Oriented Architectures allow the development of an architecture that consists of several - even spatial - dispersed hard- and software. This approach guarantees a fail safe architecture and paves the way for Ubiquitous Computing and Spatial Data Infrastructures. In addition, spatial data can be gathered and sent out in real-time from/to mobile devices utilizing Location based Services. The dissertation proves, that a specific spatial problem can be solved using mathematical modeling and an appropriate optimization technique. In this work Adaptive Large Neighborhood Search is used. By the fact that the optimization is enveloped in a Spatial Decision Support System, decision makers are involved in the decision process. On the one hand they are able to alter model parameters which alter the optimization process at hand, and on the other hand they can choose one decision alternative recommended by the system. This dissertation shows - based on example of Wood Supply Chain management - that real-time spatial optimization is able to successfully "solve" a spatial problem when coupled with a Spatial Decision Support System.
Table of Contents
1 Introduction
1.1 Supply Chain Management
1.2 Wood Supply Chain
1.3 Spatial Optimization
1.4 Problems Addressed – Hypothesis
1.5 Relevant Literature
1.6 Organization of the thesis
I Theoretical Part
2 Geographic Information Science and Technology
2.1 Standardization in GIS
2.1.1 Interoperability
2.1.2 Standards
2.1.3 De facto Standards
2.1.4 De jure Standards
2.2 Service Oriented Architectures in GIS
2.3 Location Based Services
2.4 Conclusion
3 Graph Theory and Optimization Techniques
3.1 Graph Theory
3.1.1 Graphs
3.1.2 Paths, Walks and Connected Graphs
3.1.3 Digraphs
3.2 Vehicle Routing Problems
3.2.1 VRP with Time Windows
3.2.2 VRP with Backhauls
3.2.3 VRP with Pickup and Delivery
3.2.4 VRP with Pickup and Delivery and Time Windows
3.3 Exact Optimization Techniques
3.3.1 Branch and Bound
3.3.2 Branch and Cut
3.4 Heuristical Optimization Techniques
3.4.1 Construction Methods
3.4.2 Local Search Methods
3.4.3 Simulated Annealing
3.4.4 Adaptive Large Neighborhood Search
3.5 Rolling Schedule approach
3.6 Conclusion
4 Spatial Decision Support Systems
4.1 Definition, Components and brief history of SDSS
4.2 SDSS and GISc
4.3 Integration of Optimization Techniques
4.4 Conclusion
II Application and Results
5 System Architecture and Implementation of WSC Prototype
5.1 Overview of System Architecture
5.2 Database Management System
5.3 Tracking Engine
5.4 Web Mapping Engine
5.5 Decision Engine
5.5.1 Rolling Schedule for WSC optimization
5.5.2 Graph theory for WSC optimization
5.5.3 Object oriented design for handling of optimization data
5.5.4 Retrieval of an initial solution
5.5.5 ALNS for WSC optimization
5.6 Mobile Devices
5.6.1 Mobile Device for vehicles
5.6.2 Mobile Device for forest enterprise
5.7 Desktop System
5.8 Conclusion
6 Setup and Results
6.1 Experiment Setup
6.1.1 Hardware
6.1.2 Software
6.1.3 Spatial data and area of interest
6.1.4 Problem instances
6.2 Results
6.2.1 Work Expenditure
6.2.2 Results of ALNS applied to WSC
6.3 Conclusion
7 Discussion and Outlook
7.1 Evaluation of Alternative Architectures
7.1.1 User requirements
7.1.2 Quality Evaluation of System Architectures
7.1.3 Possible system architecture designs
7.1.4 Evaluation of System Architectures
7.2 Discussion of Results of ALNS applied to WSC
7.3 Implications of real-time spatial optimization on related fields
7.4 Future aspects
7.5 Concluding Remarks
Objectives and Research Themes
The primary objective of this thesis is to develop a real-time spatial optimization system for Wood Supply Chain (WSC) management. By combining Geographic Information Science and Technology (GIS&T) and Operations Research (OR) within a Spatial Decision Support System (SDSS), the research seeks to optimize the allocation of timber to saw mills and the routing of vehicles to maximize the overall profit of the supply chain.
- Application of real-time spatial optimization to the Wood Supply Chain (WSC).
- Utilization of Service Oriented Architectures (SOA) and international standards (ISO/OGC) to eliminate media disruptions in supply chain operations.
- Development of a functional WSC prototype based on Adaptive Large Neighborhood Search (ALNS) for solving routing problems.
- Integration of mobile devices and Location Based Services (LBS) for real-time tracking and dispatching.
- Evaluation of system architectures and validation through test instances.
Excerpt from the Book
1.1 Supply Chain Management
In the last decades Supply Chain Management (SCM) emerged as a buzzword in business operations planning. Generally speaking, this concept does assume that stakeholders are not isolated islands in the market. If the market is considered as an ocean and the participating enterprises as islands that get and deliver something from/to the ocean, the market would not work as well as it does. If nobody in the ocean knows anything about the others needs, it would be hard to develop and produce goods that are really demanded. Thus, a great amount of resources is wasted due to the lack of target-oriented production and development.
The definition of the term SCM comprises the linkage between all stakeholders during the lifecycle of a product (Wannenwetsch, 2005). Generally speaking, the Supply Chain (SC) starts with the production of the raw materials and ends at the consumer. Thus, all participating institutions, enterprises or people are part of the SC (see figure 1.1). Based on this definition, also the transportation process itself is part of the SC.
Supply Chain Planning (SCP) enables every company to plan within the SC based on the market situation. Thus several processes have to be included in the planning process, based on Wannenwetsch (2005):
Summary of Chapters
1 Introduction: This chapter introduces the concept of Supply Chain Management and the specific challenges of the Wood Supply Chain, establishing the research hypothesis and setting the scope for the thesis.
2 Geographic Information Science and Technology: This chapter provides the theoretical foundation for GIS and standardization, focusing on interoperability, Service Oriented Architectures (SOA), and Location Based Services (LBS) relevant to real-time spatial data.
3 Graph Theory and Optimization Techniques: This chapter covers fundamental graph theory and various optimization methodologies, specifically detailing Vehicle Routing Problems (VRP), exact techniques, and heuristics like ALNS and the Rolling Schedule Approach.
4 Spatial Decision Support Systems: This chapter elaborates on the components and history of SDSS, discussing how GIS and mathematical models are integrated to support complex decision-making processes.
5 System Architecture and Implementation of WSC Prototype: This chapter details the technical design and implementation of the WSC prototype, describing the modules such as the Decision Engine and the Tracking Engine within a service-oriented framework.
6 Setup and Results: This chapter presents the experimental setup, including hardware and software configurations, and documents the results of applying ALNS to various timber supply test instances.
7 Discussion and Outlook: This chapter evaluates alternative system architectures, discusses the implications of real-time spatial optimization for related fields, and provides concluding remarks regarding the potential for future developments in UbiCom.
Keywords
Real-time spatial optimization, Wood Supply Chain, GIS, Operations Research, SDSS, VRPPDTW, ALNS, Service Oriented Architecture, LBS, Supply Chain Management, Heuristics, Rolling Schedule, Metaheuristics, Logistics, Timber.
Frequently Asked Questions
What is the fundamental focus of this thesis?
The thesis focuses on applying real-time spatial optimization to Wood Supply Chain (WSC) management to improve profitability by dynamically allocating timber to saw mills and optimizing vehicle routing.
What are the primary fields of research involved?
The research is interdisciplinary, combining Geographic Information Science and Technology (GIS&T), Operations Research (OR), Computer Science, and Spatial Decision Support Systems (SDSS).
What is the central research question?
The research investigates how real-time spatial optimization, implemented within a Spatial Decision Support System using GIS and OR techniques, can enhance the profitability of the Wood Supply Chain.
Which specific optimization methodology is used?
The thesis utilizes the Adaptive Large Neighborhood Search (ALNS) metaheuristic to solve the Vehicle Routing Problem with Pickup, Delivery, and Time Windows (VRPPDTW), which is the mathematical model adopted for the WSC.
How is the prototype structured?
The system follows a Service Oriented Architecture (SOA), composed of independent modules (Decision Engine, Tracking Engine, Web Mapping Engine) connected via standardized services to ensure interoperability and real-time data flow.
Which keywords best describe the work?
Key terms include Real-time spatial optimization, Wood Supply Chain, GIS, ALNS, VRP, SOA, and Spatial Decision Support Systems.
Why is the "Rolling Schedule Approach" important for the WSC?
It allows the system to cope with uncertainty by breaking the planning process into shorter intervals, optimizing for the near future where data is accurate, rather than attempting global optimization based on unreliable long-term forecasts.
How does the system utilize mobile devices?
Mobile devices on vehicles and in forest enterprises facilitate real-time data collection (tracking, pile status) and provide guidance information, bridging the gap between field operations and the logistics center.
What was the result of applying ALNS to the provided test instances?
The results showed significant objective value increases for several test runs, demonstrating that the prototype successfully optimizes the wood supply chain despite the computational complexity of the problem.
How does the thesis address the "media breaks" in traditional supply chains?
By implementing a centralized database and standardized web services (SOA), the system ensures that information flows consistently between all stakeholders, eliminating manual, inaccurate, or fragmented data handling.
- Quote paper
- Johannes Scholz (Author), 2010, Real-Time Spatial Optimization, Munich, GRIN Verlag, https://www.hausarbeiten.de/document/165048
