Scenario modelling to support industry strategic planning and decision making

The Pastoral Properties Futures Simulator (PPFS) is a dynamic systems model, developed within a participatory action research partnership with the pastoral industry of Australia's Northern Territory. The model was purpose-built to support the industry's strategic planning capacity in the face of environmental, market and institutional uncertainty. The mediated modelling process sought to maximise social learning of industry stakeholders. Simulations were conducted using scenarios representing combinations of climatic, market, institutional and technological assumptions. Stochastic parameters included rainfall and product prices. Economic and environmental performance of model farms, including greenhouse gas emissions, were estimated. A critical evaluation of the tool finds the PPFS fit for purpose. However, limitations include lack of output validation, small number of scenarios and simplistic treatment of environmental impact dimensions. With further development, the PPFS can provide a platform (a) to assist with industry planning across the whole of Northern Australia and beyond, and (b) for policy analysis and development in the context of the Australian pastoral industry.     

Agent-based management of support systems for distributed brainstorming

In this paper, we describe the design and the management of an agent-based system that supports distributed brainstorming activities. The support system is a highly coordinated IoT application composed of many locally installed interface devices, multimedia communication functions, and cloud functions that process application logic and store meeting data. The system is designed to support a variety of brainstorming sessions, so its functionalities must be modifiable and enable the system to be adapted to different environments and user requirements without any loss of performance. System accessibility should be also ensured from any location for any user. These constraints require a flexible and usable support system.We further discuss the aspects of flexibility and usability that are important in a support system for distributed brainstorming, from which we propose a conceptual schema for flexible and usable support systems. To realize this schema, we present a resource-oriented architecture that can modify the brainstorming support system’s structure and functions. Flexibility is achieved thanks to an agent-based system that manages resources and operates on them according to users’ requests.We also describe the system architecture, which is organized around a set of channels dedicated to different services proposed to the users. We present in detail a video channel that ensures user awareness during synchronized activities. We then conduct several experiments verifying the usability of important channels in the architecture and present the results of these experiments.Finally, we discuss experimental scenarios that show how the system owes its adaptability to management based on an agent organization that supports distributed brainstorming and other activities.     

A Web-based process planning optimization system for distributed design

In this paper, a process planning module, which can optimize the selection of machining resources, determination of set-up plans and sequencing of machining operations to achieve optimized process plans, has been wrapped as services and deployed in the Internet to support distributed design and manufacturing analysis. The module includes four intelligent approaches, and a Tabu search-based approach is explained in this paper to illustrate the optimization process. A Web-based prototype system has been setup for users to carry out visualization-based manipulations and process planning of design models by invoking the services remotely. The Web-based system has been integrated with a distributed feature-based design system, and the latter can generate design models and re-represent them in an XML representation based on VRML and attributes of features to provide the input of the former. Through effective utilization of the Web and Java technologies, this system is independent of the operating system, scalable and service-oriented, and can be used by a geographically distributed design team to organize concurrent engineering design activities effectively.     

Application of genetic algorithm to computer-aided process planning in distributed manufacturing environments

In a distributed manufacturing environment, factories possessing various machines and tools at different geographical locations are often combined to achieve the highest production efficiency. When jobs requiring several operations are received, feasible process plans are produced by those factories available. These process plans may vary due to different resource constraints. Therefore, obtaining an optimal or near-optimal process plan becomes important. This paper presents a genetic algorithm (GA), which, according to prescribed criteria such as minimizing processing time, could swiftly search for the optimal process plan for a single manufacturing system as well as distributed manufacturing systems. By applying the GA, the computer-aided process planning (CAPP) system can generate optimal or near-optimal process plans based on the criterion chosen. Case studies are included to demonstrate the feasibility and robustness of the approach. The main contribution of this work lies with the application of GA to CAPP in both a single and distributed manufacturing system. It is shown from the case study that the approach is comparative or better than the conventional single-factory CAPP.     

一种基于分布式强化学习的多智能体协调方法

多智能体系统研究的重点在于使功能独立的智能体通过协商、协调和协作,完成复杂的控制任务或解决复杂的问题。通过对分布式强化学习算法的研究和分析,提出了一种多智能体协调方法,协调级将复杂的系统任务进行分解,协调智能体利用中央强化学习进行子任务的分配,行为级中的任务智能体接受各自的子任务,利用独立强化学习分别选择有效的行为,协作完成系统任务。通过在Robot Soccer仿真比赛中的应用和实验,说明了基于分布式强化学习的多智能体协调方法的效果优于传统的强化学习。     

Using relaxations to improve search in distributed constraint optimisation

Densely connected distributed constraint optimisation problems (DisCOP) can be difficult to solve optimally, but finding good lower bounds on constraint costs can help to speed up search. We show how good lower bounds can be found by solving relaxed problems obtained by removing inter-agent constraints. We present modifications to the Adopt DisCOP algorithm that allow an arbitrary number of relaxations to be performed prior to solving the original problem. We identify useful relaxations based on the solving structure used by Adopt, and demonstrate that when these relaxations are incorporated as part of the search it can lead to significant performance improvements. In particular, where agents have significant local constraint costs, we achieve over an order of magnitude reduction in messages exchanged between agents. Finally, we identify cases where such relaxation techniques produce less consistent benefits.     

Conceptualizing a shared language subsystem for distributed decision support systems

Researchers have only recently begun to develop formal architectures for support systems that capitalize on advancements in distributed technologies. Conceptual models have been proposed for distributed versions of both DSS and GDSS. To date, however, these developments have largely been limited to techniques employing shared global data spaces. Research on other approaches to distributed support systems development, including that based on the shared language techniques of distributed artificial intelligence, is still lagging. To introduce this research area, this paper reviews architectural principles for distributed decision support and develops the notion of a shared language as an extension of the DSS language subsystem.     

e-CAPP: A distributed collaborative system for internet-based process planning

Advances in distributed technologies have enabled engineers to communicate more effectively, collaborate, obtain, and exchange a wide range of design resources during product development. Shared internet-based virtual environments allow experts in remote locations to analyze a virtual prototype, together and simultaneously in centers in which the product is being developed.This paper presents a system for distributed and collaborative environment which could assist manufacturing enterprises and experts in discussing, suggesting, evaluating and selecting best process plans for family of manufacturing parts. The represented e-CAPP system enables the implementation of expert knowledge in an appropriate knowledge repository. The knowledge from this repository is integrated into intra-company CAPP systems and used while generating process plans for new products. The proposed internet-based collaborative environment, dedicated to distributed process planning, is yet another step in the direction advancing of distributed manufacturing.     

Formal models and algorithms for decentralized decision making under uncertainty

Over the last 5 years, the AI community has shown considerable interest in decentralized control of multiple decision makers or “agents” under uncertainty. This problem arises in many application domains, such as multi-robot coordination, manufacturing, information gathering, and load balancing. Such problems must be treated as decentralized decision problems because each agent may have different partial information about the other agents and about the state of the world. It has been shown that these problems are significantly harder than their centralized counterparts, requiring new formal models and algorithms to be developed. Rapid progress in recent years has produced a number of different frameworks, complexity results, and planning algorithms. The objectives of this paper are to provide a comprehensive overview of these results, to compare and contrast the existing frameworks, and to provide a deeper understanding of their relationships with one another, their strengths, and their weaknesses. While we focus on cooperative systems, we do point out important connections with game-theoretic approaches. We analyze five different formal frameworks, three different optimal algorithms, as well as a series of approximation techniques. The paper provides interesting insights into the structure of decentralized problems, the expressiveness of the various models, and the relative advantages and limitations of the different solution techniques. A better understanding of these issues will facilitate further progress in the field and help resolve several open problems that we identify. This work was done while S. Seuken was a graduate student in the Computer Science Department of the University of Massachusetts, Amherst.     

多agent规划综述*

多Agent规划是智能规划和多Agent系统的交叉领域,随着智能规划领域研究范围的不断扩展和多Agent系统领域研究内容的不断深入,多Agent规划受到了越来越多的关注。有鉴于此,本文对多Agent规划的概念和主要方法进行全面综述。具体内容包括智能规划和多Agent系统的背景介绍、多Agent规划的各种形式化描述方式以及基于规划修复、合并或者马尔可夫决策过程的的分布式规划方法。最后,还给出了多Agent规划的发展趋势。     

A visualization tool to support decision making in environmental and biological planning

Large-scale ecosystem management involves consideration of many factors for informed decision making. The EverVIEW Data Viewer is a cross-platform desktop decision support tool to help decision makers compare simulation model outputs from competing plans for restoring Florida's Greater Everglades. The integration of NetCDF metadata conventions into EverVIEW allows end-users from multiple institutions within and beyond the Everglades restoration community to share information and tools. Our development process incorporates continuous interaction with targeted end-users for increased likelihood of adoption. One of EverVIEW's signature features is side-by-side map panels, which can be used to simultaneously compare species or habitat impacts from alternative restoration plans. Other features include examination of potential restoration plan impacts across multiple geographic or tabular displays, and animation through time. As a result of an iterative, standards-driven approach, EverVIEW is relevant to large-scale planning beyond Florida, and is used in multiple biological planning efforts in the United States.     

Integrated yet distributed operations planning approach: A next generation manufacturing planning system

Present paper envisages the need for an innovative operations planning system to handle the challenges and opportunities offered by next industrial revolution called Industry 4.0 or smart manufacturing. In specific, to embrace the increasing level of automation in manufacturing industries, the obligation of joint consideration of multiple operations functions is realized. On the other hand, quick response to dynamic conditions created by machine failures, change in demand, uncertainty in supply, etc., is important in captivating the advantages of the digitization in industries. Easing out the computational complexity, imposed by the integration of multiple functions, therefore, becomes an important aspect of next generation manufacturing planning systems. Consequently, in this paper, an agent-based approach is engineered around the opportunities offered by modern digital factory viz., intelligence at the shop-floor and ubiquity of wireless communications. While intelligence at shop-floor allows distributing the decision-making tasks to various functional agents, the communication among the agents makes it feasible to incite integrated view through the coordination agent. The approach is demonstrated for a representative industrial environment of an automotive plant. Further, comparison over conventional approaches, computational comparison, effect of degree of integration, and performance of the approach under dynamic conditions are investigated. Finally, the approach is comprehensively evaluated to analyze its robustness and implications in various manufacturing settings. This extensive investigation shows that the proposed operations planning system has capability to apprehend the benefits from next generation intelligent factory.     

Spillover Algorithm: A decentralised coordination approach for multi-robot production planning in open shared factories

Open and shared manufacturing factories typically dispose of a limited number of industrial robots and/or other production resources that should be properly allocated to tasks in time for an effective and efficient system performance. In particular, we deal with the dynamic capacitated production planning problem with sequence independent setup costs where quantities of products to manufacture need to be determined at consecutive periods within a given time horizon and products can be anticipated or back-ordered related to the demand period. We consider a decentralised multi-agent variant of this problem in an open factory setting with multiple owners of robots as well as different owners of the items to be produced, both considered self-interested and individually rational. Existing solution approaches to the classic constrained lot-sizing problem are centralised exact methods that require sharing of global knowledge of all the participants’ private and sensitive information and are not applicable in the described multi-agent context. Therefore, we propose a computationally efficient decentralised approach based on the spillover effect that solves this NP-hard problem by distributing decisions in an intrinsically decentralised multi-agent system environment while protecting private and sensitive information. To the best of our knowledge, this is the first decentralised algorithm for the solution of the studied problem in intrinsically decentralised environments where production resources and/or products are owned by multiple stakeholders with possibly conflicting objectives. To show its efficiency, the performance of the Spillover Algorithm is benchmarked against state-of-the-art commercial solver CPLEX 12.8.     

An agent-based approach to knowledge management in distributed design

The collaboration process among individuals with heterogeneous skills in a distributed virtual environment represents a crucial element of the extended enterprise. In order to achieve global optima in design, there is an increasing need for design teams to establish and maintain cooperative work through effective communication, co-location, coordination and collaboration at the knowledge level. Because of the distributed nature of users and information resources involved in the design process, the need for appropriate knowledge management tools is imperative. This paper proposes an agent-based architecture to support multi-disciplinary design teams that cooperate in a distributed design environment (DDE). Using ontologies and multi-agent systems (MAS), the proposed framework aims to optimise design process operation and management by supporting the dialogue among distributed design actors. Received: February 2005/Accepted: January 2006     

Coordination and control for energy distribution in distributed grid networks: Theory and application to power dispatch problem

This paper presents a new framework considering decentralized energy coordination and generation, and flow control for supply–demand balance in distributed grid networks. Consensus schemes using only local information are employed to produce energy coordination, generation, and flow control signals. For the supply–demand balance, it is required to determine the amount of energy needed at each distributed resource. Also, due to the different generation capacities of each energy resource, coordination of energy flows among distributed energy resources is essentially required. Thus, this paper proposes a new framework which gives decentralized energy coordination scheme, generation, and flow control method considering these constraints based on distributed consensus algorithms. The proposed framework in this paper can be nicely utilized in energy dispatch or energy flow scheduling. Furthermore, it can be applied to various engineering problems including water irrigation systems, traffic networks, and building automation systems since it deals with attributed distribution and resource allocation in large scale distributed systems. Through illustrative examples, the effectiveness of the proposed approaches is illustrated. A possible application to power dispatch problem in the IEEE-14bus is also addressed for more detailed and realistic evaluation.     

New perspectives in decision support for port planning

The Decision-Support-System (DSS) to be presented within this paper is based on research done at the department of Computer Science of the Technical University of Berlin. Among the research conducted at the subdepartment for Applied Computer Science, the project “Determination of the Capacity of Ports” was one point of focus in the years 1980–1985. Findings of this research work as well as practical experiences helped in underpinning the knowledge of the decision processes in seaports; all this contributed to the development of a practice-oriented DSS. All parts of the system, including the data base, were programmed in SIMULA and implemented on the IBM 4381-2 of the Technical University of Berlin. Before the system is applied in an actual problem situation, all parts of the system, particularly the integrated simulation model, will have to be validated. Not until this validation has been done, will the decision makers have sufficient confidence regarding the decision support of this system. This validation is still being worked on. Further research is focused on the application of the DSS for similar decision problems in other traffic and transportation systems.     

多Agent系统合作与协调机制研究综述

多Agent间的合作和协调机制是多Agent系统理论的核心问题之一。本文按照协调在合作过程中发生的位置进行分类讨论,结合实例阐述了经典的多Agent系统合作与协调机制的主要原理、适用范围,旨在分析总结现有的多Agent合作和协调机制,为未来的开放、异构、动态环境下多Agent合作与协调机制的发展趋势和进一步研究的方向提供参考。     

Towards a generic distributed and collaborative digital manufacturing

A framework for distributed manufacturing is proposed to facilitate collaborative product development and production among geographically distributed functional agents using digitalized information. Considering the complexity of products created in a distributed manufacturing scenario, it often requires close collaborations among a number of facilities. In this research work, various functional agents, such as the manufacturability evaluation agent (MEA), manufacturing resource agent (MRA), process-planning agent (PPA), manufacturing scheduling agent (MSA), shop floor agent (SFA), fault diagnosis agent (FDA), etc., can interact coherently for distributed manufacturing. With specific agents having unique functionalities, a manufacturing managing agent (MMA) acts as the centre of this distributed manufacturing system. The MMA agent assists the specific agents’ to work seamlessly and also to collaborate closely with the participating agents. In this way, the production cycle of a part can be optimized from product design to final manufacturing since all the production procedures are considered logically and every procedure is correlated. The agent language based on the knowledge query manipulation language (KQML) includes many pre-defined performatives that ease the participating agents to carry out their tasks intelligently by interpreting commands from one another. Additionally, to ensure the adaptiveness and upgradeability of the system, the internal structure of each functional agent that is based on JATLite is modularized into several components, including a communication interface, central work engine, knowledge base pool, and input/output modifier for possible future methodology enhancements.     

A multi-agent system for minimizing energy costs in cement production

A cement production planning system is expected to minimize energy costs. Further, such a system needs to be as autonomous as possible to decrease time loss during the communication between related departments of the plant. Hence, in this paper, we present a multi-agent system (MAS) in which software agents work collaboratively in order to assist production, planning and sales departments of a cement plant for the generation of cost-effective cement production plans. Implemented system was deployed and actively used inside one of the plants of a leading cement company in Turkey. Evaluation result shows that the utilization of the proposed system caused a significant energy cost saving. Moreover, workers in the planning department of the cement plant saved approximately 75% of their working hour by using the system. Total workload of the employees (including all departments) decreased to its half.