Emergent Intelligence in Smart Ecosystems: Conflicts Resolution by Reaching Consensus in Resource Management

Abstract

A new emergent intelligence approach to the design of smart ecosystems, based on the complexity science principles, is introduced and discussed. The smart ecosystem for resource management is defined as a system of autonomous decision-making multi-agent systems capable to allocate resources, plan orders for resources, and to optimize, coordinate, monitor, and control the execution of plans in real time. The emergent intelligence enables software agents to collectively resolve conflicts arising in resource management decisions by reaching a consensus through a process of detecting conflicts and negotiation for finding trade-offs. The key feature of the proposed approach is the ontological model of the enterprise and the method of collective decision-making by software agents that compete or cooperate with each other on the virtual market of the digital ecosystem. Emergent intelligent systems do not require extensive training using a large quantity of data, like conventional artificial intelligence/machine learning systems. The developed model, method, and tool were applied for managing the resources of a factory workshop, a group of small satellites, and some other applications. A comparison of the developed and traditional tools is given. The new metric for measuring the adaptability of emergent intelligence is introduced. The performance of the new model and method are validated by constructing and evaluating large-scale resource management solutions for commercial clients. As demonstrated, the essential benefit is the high adaptability and efficiency of the resource management systems when operating under complex and dynamic market conditions.

Keywords

• emergent intelligence
• smart ecosystems
• decision making
• autonomous systems
• complex adaptive systems
• self-organization