Agent-based decision-support framework for water supply infrastructure rehabilitation and development

The broad platform of our research addresses the possibilities offered by the concept of complete agents as a theoretical basis for designing computational architectures of use in modelling intelligent behaviours. One area that we are investigating is the application of the computational metaphor of agents to decision-support systems. Here we present the current state of our progress in developing a framework of use in making decisions about water supply infrastructure rehabilitation and development. The framework currently supports database reclamation, data warehousing, water mains pipe-failure prediction and strategic overview information based on customer complaints and chemical analysis of supplied water. Further aspects of the framework have been developed for other problem domains. This ongoing research draws on methodologies from work on agents, knowledge engineering, data-mining and knowledge discovery.