The breadth of agent based applications in industry

The field of software agents has quickly gained ground from being an interesting and novel research area to now developing practical applications. Originally the field brought together aspects of distributed AI, machine learning and autonomous systems to make an eclectic and stimulating field of research. Who would have thought that this would lead to applications as diverse as workflow management, personal assistants and electronic commerce? This augmentation of autonomy, learning, and distribution leads to a number of positive benefits for systems including these component parts. Distributed systems give the benefits of robustness, localisation and specialisation of applications by allowing them to be tailored to the local environment. They also make load balancing over networks simpler. Autonomy and autonomous systems give benefit by being very robust and lead to “fire and forget” agents. Machine learning has enabled agents to get personal and adapt to differing needs of users or systems over time. They are therefore less brittle which potentially gives them a longer lifetime