Machine Agents Research Articles

New media in situ: the re-socialisation of public space

Five recently created artworks described in this paper constitute experimental research frameworks for revitalising strategies of representation by addressing the issues of inhabitation and the sensorial in virtual environments. Their immersive architectures provide a modular interactive cinema with digitally augmented full body engagement and offer new opportunities for rendering the experience of culture and heritage. Place-Hampi, Hampi-Live, The Eye of Nagaur, T_Visionarium and UNMAKEABLELOVE are set inside four originally conceived immersive display systems: advanced visualisation and interaction environment (AVIE), Place, EoN and Re-Actor. Each of these artworks elucidates the singular qualities of physically located and socially shared new media experiences, which are both complimentary to and distinct from those experiences generated in cyberspace. The works demonstrate singular solutions to current challenges in the fields of cultural heritage, the navigation of massive multi-modal databases, interactive cinema and the evolving relationship between human and machine agents. The text frames these works in terms of cultural imaginary, co-evolutionary narrative, prosthetic vision, re-combinatory narrative and complicit agency.

Intelligent and collaborative Multi-Agent System to generate and schedule production orders

The authors present a Multi-Agent System for constructing and releasing production orders. In a manufacturing enterprise, the generation of production orders consists in a set of coordinated tasks among departments. This has been achieved traditionally as a module of the Production Activity Control (PAC) system. However, classic PAC modules lack collaborative techniques and intelligent behaviour. Moreover, in real-life situations experienced planners take over traditional PAC systems, since the range of possibilities to actually build production orders increases exponentially. To contribute to production planning, we present an intelligent and collaborative Multi-Agent System (MAS), having coordinated two forms to emulate intelligence. The learning capability is achieved by means of a Feed-forward Artificial Neural Network (FANN) with the back-propagation algorithm. The FANN is embedded within a machine agent whose objective is to obtain the appropriate machine in order to comply with requirements coming from the sales department. Also, an expert system is provided to a tool agent, which in turn is in charge of inferring the right tooling. The MAS also consists of a coordinator and a spy. The coordinator agent has the responsibility to control the flow of messages among the agents, whereas the spy agent is constantly reading the Enterprise Information System. Finally, a scheduler agent schedules the production orders. The resultant MAS improves the current form to plan production in a factory dedicated to produce labels.

Agent-based FMS control

Future manufacturing systems will be integrated into the networks of distributed resources, and at the same time, such systems will be capable of processing both knowledge and material. It will probably be required that manufacturing systems be agile, flexible, and fault-tolerant. Petri nets (PN) and object-oriented design (OOD) are used together in order to develop the integrated agent-based FMS control system. The flexible manufacturing system (FMS) consists of machines, workstations, and automated material handling system, distributed buffer storage sites and computer-based supervisory control, all which can be modeled as an agent in OOD with PN. This paper introduces the design of an agent-based FMS control system through PNs and evaluates the performance using timed placed Petri nets (TPPN). In order to do so, the agent control design, FMS structure has been evaluated in detail and the agent definitions have been submitted. The system includes the sharing and distribution of tasks among agents and the mentioned structure has been simulated by TPPN. The simulation procedure has been realized through Petri Net 2.0—MATLAB Demo Program [Mahulea CF, Motcovschi MH, Pastravanu O. Department of Automatic Control Industrial Informatics, Technical University “Gh. Asachi” of Iasi, Blvd., Mangeron 53A, 6600 Iasi, Romania, 〈 http://www.ac.tuiasi.ro/pntool,pntool@ac.tuiasi.ro〉, 2004.]. Each case is modeled, and then the agent's machine processing time is considered in this program. As for the evaluation of the study, the system performance is assessed through the waiting time of the parts in queue and the task distributions.

A multi-agent system to construct production orders by employing an expert system and a neural network

The authors describe the implementation of a multi-agent system, whose goal is to enhance production planning i.e. to improve the construction of production orders. This task has been carried out traditionally by the module known as production activity control (PAC). However, classic PAC systems lack adaptive techniques and intelligent behaviour. As a result they are mostly unfit to handle the NP Hard combinatorial problem underlying the construction of right production orders. To overcome this situation, we illustrate how an intelligent and collaborative multi-agent system (MAS) obtains a correct production order by coordinating two different techniques to emulate intelligence. One technique is performed by a feed-forward neural network (FANN), which is embedded in a machine agent, the objective being to determine the appropriate machine in order to fulfil clients’ requirements. Also, an expert system is provided to a tool agent, which in turn is in charge of inferring the right tooling. The entire MAS consists of a coordinator, a spy, and a scheduler. The coordinator agent has the responsibility to control the flow of messages among the agents, whereas the spy agent is constantly reading the Enterprise Information System. The scheduler agent programs the production orders. We achieve a realistic MAS that fully automates the construction and dispatch of valid production orders in a factory dedicated to produce labels.

Implementation of a supervised learning technique in a multi-agent system for building production orders

The authors describe the implementation of a supervised learning algorithm within a multi-agent system, whose general objective is to build production orders. Although this task has been carried out traditionally by the production management system, the classic approach lacks adaptive techniques and intelligent behavior. It is acknowledged that the combinatorial problem underlying the construction of production orders belongs to the NP hard complexity class. Therefore, flexible computational solutions are needed. We claim that by using intelligence and collaboration in a multi-agent system (MAS), a correct solution is reached more efficiently. Intelligence is emulated by both learning and decision-making, achieved through a feed-forward artificial neural network (FANN). The FANN is embedded in a machine agent, which determines the appropriate machine to manufacture the product. Collaboration is obtained by employing a sound protocol based on FIPA-ACL messages. We illustrate the approach by designing and implementing a MAS, which is already in use in a company that produces labels.

Product and machine agents for an autonomous assembly production system

In this paper, product parts and machine agents are proposed for the configuration of an autonomous flexible assembly system. Product parts are divided into two types. One is the main part that is a base of the assembled product and the other is for sub parts that are installed on the main part. These parts and assembly machines are implemented as software agents in a virtual assembly production line. The main part agent takes the initiative in this system. When a new main part is put on to the assembly line, the main part agent with the product model and sub part agents are dynamically generated. The main part agent plans the assembly process by itself and selects assembly machines that execute the assembly operations by communicating with assembly machine agents and other main part agents. The part agents are deleted when a product is completed. The assembly machine agent has a machine model of itself and can determine which operations can be processed by the machine itself, The product model and machine model are described using XML. Trial implementation of main part agents, sub parts agents and assembly machine agents were done as was an assembly production simulation.

A Multiagent-Based Decision-Making System for Semiconductor Wafer Fabrication With Hard Temporal Constraints

This paper presents a decision-making system for semiconductor wafer fabrication facilities, or wafer fabs, with hard interoperation temporal constraints. The decision-making system is developed based on a multiagent architecture that is composed of scheduling agents, workcell agents, machine agents, and product agents. The decision-making problem is to allocate lots into each workcell to satisfy both logical and temporal constraints. A dynamic planning-based approach is adopted for the decision-making mechanism so that the dynamic behaviors of the wafer fab such as aperiodic lot arrivals and reconfiguration can be taken into consideration. The scheduling agents compute quasi-optimal schedules through a bidding mechanism with the workcell agents. The proposed decision-making mechanism uses a concept of temporal constraint sets to obtain a feasible schedule in polynomial steps. The computational complexity of the decision-making mechanism is proven to be, where is the number of operations of a lot and is the cardinality of the temporal constraint set.

4104 部品と機械のエージェント化による自律型組立システムの構成(OS4-1 循環生産システム・新しい生産システムアーキテクチャ)

4104 部品と機械のエージェント化による自律型組立システムの構成(OS4-1 循環生産システム・新しい生産システムアーキテクチャ)

Development of a hybrid negotiation scheme for multi-agent manufacturing systems

A hybrid inter-agent negotiation mechanism based on currency and a pre-emption control scheme is proposed to improve the performance of multi-agent manufacturing systems. The multi-agent system considered consists mainly of four types of agents: machine, clone, part and mediator. The machine agent controls the scheduling and the execution of a task. The clone agent aims to maximize the utilization rate by attracting relevant work to the machine. The part agent communicates with the machine agent or clone agent to acquire necessary production resources in order to get the required processing done, and the mediator agent contains the status of the part that will be processed by the subcontracting machine agent. The primary objective is to design decentralized control protocols for discrete part manufacturing systems to enhance the efficiency of the system and to allocate dynamically the resources to critical jobs based on the dynamic search tree. This research incorporates both the currency and the pre-emption schemes within a common framework. Currency functions are used to help the agents meet their individual objectives, whereas the pre-emption scheme is used to expedite the processing of parts based on their due dates. A dynamic search algorithm for the best route selection of different operations based on the job completion time is also proposed and it is implemented on a small manufacturing unit.

Multiagent-based modeling for re-entrant manufacturing system

Re-entrant flow manufacturing system is a typical mix of different process types, sequence-dependent setup times, very expensive equipment, and re-entrant flows. A multiagent-based modeling technology for re-entrant manufacturing system was addressed. The model of Agent-based Colored Timed Petri-Net (ACTPN) is constructed, which consists of the material flow net model and the interactive protocol model. In the material flow net, the interior behavior of machine agent was encapsulated and it was separated into the system layer, machine group layer and machine agent layer to form a hierarchical multiagent-based model, which reduced the complexity and improved the reusability of the model. In the view of modeling methodology of the interactive protocol net, the transforming rule from Agent Unified Modeling Language (AUML) to ACTPN was given so as to map the AUML model to ACTPN model. In the view of the analysis method of the interactive protocol net, it was integrated with a reachability graph to check the liveness and bound properties of interactive protocol model. A distributed discrete event simulation platform to evaluate the performance of the multiagent-based re-entrant manufacturing system was built. An illustrative example of a 6-inch semiconductor fabrication line in Shanghai demonstrates the effectiveness of the approach proposed.

Bidding-based multi-agent system for integrated process planning and scheduling: a data-mining and hybrid tabu-SA algorithm-oriented approach

This paper conceptualizes a bidding-based multi-agent system for solving integrated process-planning and scheduling problem. The proposed architecture consists of various autonomous agents capable of communicating (bidding) with each other and making decisions based on their knowledge. Moreover, in contrast to the traditional model of integrated process-planning and scheduling problem, a new paradigm has been conceptualized by considering tool cost as a dynamic quantity rather than a constant. Tool cost is assumed to comprise tool-using cost and its repairing cost. The repairing cost is considered to depend on the tool-breaking probability, which is predicted by the data-mining agent equipped with the virtues of C-fuzzy decision tree. When a job arrives at the shop floor, the component agent announces a bid for one feature at a time to all the machine agents. Among the machine agents capable of producing the first feature, one comes forward to become a “leader”, and groups other machine agents for the processing of remaining features of the job. Once all features are assigned to the appropriate machines, the leader then sends this allocation information to the optimization agent. The optimization agent finds optimal/near-optimal process plans and schedules via the hybrid tabu-SA algorithm.

Ant colony intelligence in multi-agent dynamic manufacturing scheduling

This study aims at building an efficient agent-based dynamic scheduling for real-world manufacturing systems with various products, processes, and disturbances. Ant colony intelligence (ACI) is proposed to be combined with local agent coordination so as to make autonomous agents adaptive to changing circumstances and to give rise to efficient global performance. The work here differs from other dynamic scheduling research in two areas: (1) a more generic and realistic manufacturing model with multiple product types, multiple/parallel multi-purpose machines with sequence-dependent setup constraints, and various dynamic disturbances is used, (2) ACI integrated with both machine agents and job agents to solve not only the task allocation problem, but also the task sequencing problem. The implementation of the aforementioned issues in a multi-agent system (MAS) is discussed. Simulation results show that, for most of the performance measures, a MAS integrated with well-designed ant-inspired coordination performs well compared to a MAS using dispatching rules.

Communication Strategies From High-reliability Organizations

Communication Strategies From High-reliability Organizations

PANEL ON AUTHORITY DISTRIBUTION IN HUMAN-MACHINE SYSTEMS

Even if a large amount of work has been done so far in both technological and human factors worlds, automation is still a controversial issue. Human and technological reliabilities cannot be taken separately. Human-machine systems are more intimately integrated, in particular with massive use of software as mediating spaces. The main issue is to better understand who must be in charge, verify who is effectively in charge and so on. Safe, efficient and comfortable environments and systems require enhanced analysis of responsibility allocation among the various human and machine agents. Such analysis should take into account not only the increasing distance between humans and the physical world that they try to control, but the increasing number of both artificial and human agents that mediate interactions. This panel will provide a discussion forum to advance these topics.

Evaluating the Effectiveness of a Joint Cognitive System: Metrics, Techniques, and Frameworks

An implication of Cognitive Systems Engineering is that joint cognitive systems (JCS; also known as complex socio-technical systems) need to be evaluated for its effectiveness in performing the complex cognitive work requirements. This requires using measures that go well beyond “typical” performance metrics such as the number of subtask goals achieved per person per unit of time and the corresponding simple baseline comparisons or workload assessment metrics. This JCS perspective implies that the system must be designed and evaluated from the perspective of the shift in role of the human supervisor. This imposes new types of requirements on the human operator. Previous research in CSE and our own experience has lead us to identify a set of generic JCS support requirements that apply to cognitive work by any cognitive agent or any set of cognitive agents, including teams of people and machine agents. Metrics will have to reflect such phenomena as “teamwork” or “resilience” of a JCS. This places new burdens on evaluation techniques and frameworks, since metrics should be generated from a principled approach and based on fundamental principles of interest to the designers of the JCS. An implication of the JCS perspective is that complex and cognitive systems need to be evaluated for usability, usefulness, and understandability; each of which goes well beyond raw performance. However, conceptually-grounded evaluation frameworks, corresponding operational techniques, and corresponding measures for these are limited. Therefore, in order to advance the state of the field, we have gathered a set of researchers and practitioners to present recent evaluation work to stimulate discussion.

Integrated process planning and scheduling/rescheduling—an agent-based approach

This paper presents a hybrid-based multi-agent system (MAS) for integrating process planning with scheduling/rescheduling in job shops or similar kinds of flexible manufacturing environments. The approach is called Online Hybrid Agent-based Negotiation (oHAN). It employs the affected operations rescheduling approach to reschedule the operations affected directly or indirectly by the disruptions. The proposed MAS architecture comprises local agents, which are the part and machine agents, and a supervisor agent. The local agents detect the production disturbances and reschedule the affected operations via decentralized negotiations. A comprehensive set of process plan flexibilities including flexible routings and alternative machines is considered during rescheduling. The supervisor agent coordinates and monitors the rescheduling process in order to ensure the observance of a global rescheduling objective. An online hybrid contract-net negotiation protocol (oHCNP) has been established to control the interactions and communications among agents. Two types of disturbances, machine breakdown and new part arrival, are investigated in this paper. Extensive experiments have been conducted and the results show that the proposed approach is effective in solving a large-scale rescheduling problem in a reasonable time. With the introduction of the supervisory control into the decentralized negotiations, the hybrid approach is able to provide solutions with a better global performance.

An agent-based negotiation approach to integrate process planning and scheduling

This paper presents the development of an agent-based negotiation approach to integrate process planning and scheduling (IPPS) in a job shop kind of flexible manufacturing environment. The agent-based system comprises two types of agents, part agents and machine agents, to represent parts and machines respectively. For each part, all feasible manufacturing processes and routings are recorded as alternative process plans. Similarly, alternative machines for an operation are also considered. With regard to the scheduling requirements and the alternative process plans of a part, the proposed agent-based IPPS system aims to specify the process routing and to assign the manufacturing resources effectively. To establish task allocations, the part and machine agents have to engage in bidding. Bids are evaluated in accordance with a currency function which considers an agent's multi-objectives and IPPS parameters. A negotiation protocol is developed for negotiations between the part agents and the machine agents. The protocol is modified from the contract net protocol to cater for the multiple-task and many-to-many negotiations in this paper. An agent-based framework is established to simulate the proposed IPPS approach. Experiments are conducted to evaluate the performance of the proposed system. The performance measures, including makespan and flowtime, are compared with those of a search technique based on a co-evolutionary algorithm.

반도체 웨이퍼 팹의 에이전트 기반 스케쥴링 방법

This paper proposes an agent-based scheduling method for semiconductor wafer fabrication facilities with hard inter-operation temporal constraints. The scheduling problem is to find the feasible schedules that guarantee both logical and temporal correctness. A proposed multi-agent based architecture is composed of scheduling agents, workcell agents, and machine agents. A scheduling agent computes optimal schedules through bidding mechanisms with a subset or entire set of the workcell agents. A dynamic planning-based approach is adopted for the scheduling mechanism so that the dynamic behaviors such as aperiodic job arrivals and reconfiguration can be taken into consideration.

Dynamic shopfloor scheduling in multi-agent manufacturing systems

Abstract This paper is on the development of an agent-based approach for the dynamic integration of the process planning and scheduling functions. In consideration of the alternative processes and alternative machines for the production of each part, the actual selection of the schedule and allocation of manufacturing resources is achieved through negotiation among the part and machine agents which represent the parts and manufacturing resources, respectively. The agents are to negotiate on a fictitious cost with the adoption of a currency function. Two MAS architectures are evaluated in this paper. One is a simple MAS architecture comprises part agents and machine agents only; the other one involves the addition of a supervisor agent to establish a hybrid-based MAS architecture. A hybrid contract net protocol is developed in the paper to support both types of MAS architectures. This new negotiation protocol enables multi-task many-to-many negotiations, it also incorporates global control into the decentralized negotiation. Simulation runs are conducted to compare the performance of the proposed MAS-based IPPS approaches and that of an evolutionary algorithm. It also shows that the hybrid-based MAS, with the introduction of supervisory control, is able to provide integrated process plan and job shop scheduling solutions with a better global performance.

A systemic perspective of situation awareness

PurposeMaintaining a high level of situation awareness (SA) is considered one of the most essential elements for safe and effective flight operations. In a study of accidents among major air carriers, 88 per cent of those involving human error could be attributed to problems with SA. In complex domains such as aviation, SA is inherently distributed over multiple people and groups and over human and machine agents. The purpose of this paper is to present an alternative perspective to the hegemony of the cognitive approach to SA that focuses on the systemic nature of SA.Design/methodology/approachAn alternative approach to the hegemony of the cognitive perspective of SA has been presented, that focuses on a systemic or holistic conceptualisation of SA through the application of Somerville's actor network theory (ANT). By advocating a seamless web composed of actors, the actor network approach dissolves the dichotomous relationship between humans and machines and society and technology into a non‐anthropocentric framework. This paper further develops this systemic perspective of SA through an analysis of the tragic 2002 mid‐air collision over Überlingen, Germany case study.FindingsThe application of ANT to this case study brings to light some insights with wide ranging consequences for how we think of SA and accident aetiology.Practical implicationsThe systemic perspective of SA has far‐reaching design implications with regard to complex socio‐technical systems.Originality/valueThis paper facilitates the perspective that looks at the inter‐connectedness of the heterogeneous elements characterized by the technological and non‐technological (human, social, organizational, political) elements of complex socio‐technical systems.