Please login first
Agent-based Solutions for Industrial Environments composed of Autonomous Mobile Agents, Modular Agent Platforms, and Tuple Spaces.
1  University of Bremen, Dept. of Mathematics and Computer Science, Robert Hooke Str. 5, 28359 Bremen, Germany


Future factory production and assembly environments require smart automation and are controlled by a massively increasing number of computers with sensorial feedback from machines, parts, products, and humans, consisting of a wide variety of different networked devices and software programs, which can be considered as one big use case of pervasive and cloud computing with vanishing boundaries between the computing and the environment, and with a strong focus on decentralized distributed computing and information storage. These new complex information processing architectures are composed of hierarchical network graphs, and require some kind of self-organization and adaptability to overcome single-point of failure and robustness constraints. The data acquired from machines and sensitive products or parts are growing at a fast rate, leading to a large data volume that must be handled distributed  with pre-processing, map- and reduce, and filtering algorithms.

Mobile Agents can be deployed in such large-scale and hierarchical network environments crossing barriers transparently, for example, industrial manufacturing and assembly environments, the Internet, Sensor Networks, and Cyber-Physical Systems. The networks can consist of high- and low-resource nodes ranging from generic computers to microchips, and the supported network classes range from body area networks to the Internet including any kind of sensor and ambient network. Mobile Agents can perform distributed computation in an autonomous manner.

In this work Agents are represented by mobile program code that can be modified at run-time by the Agents themselves. The presented approach enables the development of sensor clouds and smart systems of the future integrated in daily use computing environments and the Internet. Agents can migrate between different hardware and software platforms by migrating the program code of the agent, which embeds  the control state and the private data of an agent, finally encapsulated in self-initializing and self-containing code frames.

This cross-platform interoperability is ensured by a modular and scalable Agent Processing Platform. The entire information exchange and co-ordination of Agents with other Agents and the environment is performed by using a Tuple Space database, unifying the platform and architecture specific data representation. The Tuple Space is used for any kind of information exchange including program code and directory and file system services mapped on the Tuple Space. Beside architecture specific hardware and software implementations of the agent processing platform, there is a JavaScript (JS) implementation layered on the top of a distributed co-ordination and management layer including distributed file and name services. The JS platform enables the integration of Multi-Agent Systems (MAS) in Internet server and application environments (e.g., WEB browser). Agents can migrate transparently between different classes of computing devices and environments, ranging from hardware-level sensor networks (embedded in technical structures) to WEB browser applications or network servers without any required transformation.

Keywords: Industrial Agents, Adaptive Manufacturing, Holonic Systems, Reliability, Heterogeneous Networks, Internet-of-Things, Mogile Agents, Self-organizing and Self-adaptive Systems