Issue No. 02 - March/April (1999 vol. 14)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/5254.757632
<p>Agent technology is in a state of paradox. The field has never enjoyed more energy and concomitant research progress, and yet the rate of uptake of new research results in fielded systems has been glacially slow. The few agents in the real world of everyday applications generate more heat than enlightenment; most are easily confused, few collaborate except in trivial prearranged fashion, and all enjoy little freedom of movement. Significantly, the current trapped state of our agents has less to do with lack of mobility mechanisms than with their unpreparedness to work fully in the open world of cyberspace and to interoperate outside of a tightly circumscribed sphere of agent platforms and domains. The kinds of agents that we want-citizens of the wired world, equipped with stamped passports and Berlitz travelers guides explaining foreign phrases and places that allow them to hail, meet, and greet agents of any sort in the open landscape of the Internet and, if not able to team up on a project, at least able to ask intelligibly for directions-these kinds of agents, alas, exist today only in our imaginations (and, of course, in the vision sections of our research proposals).</p> <p>Actually building the sophisticated agent-based systems of the future will require research advances on at least three fronts: We must continue work on agent theory so that many currently unanswered questions about the scope and limitations of alternative approaches to agent design can be addressed. We must make agent frameworks and infrastructure powerful, interoperable, and secure enough to support robust large-scale coordinated problem-solving activity.</p> <p>Perhaps more importantly, we must develop new sorts of tools to help nonspecialists unlock the power of agent technology.</p> <p>The good news is that things are progressing well on the first two fronts. Various initiatives are beginning to provide an early preview of the faster, more reliable, and more secure versions of the next-generation Internet that our large-scale visions require. Middleware and Internet technologies and standards are now maturing to the point that agent framework developers can rely on off-the-shelf products as a ready substrate to their own work, rather than creating ad hoc alternatives from scratch. Advances in the difficult theoretical issues of dynamic agent communication, coordination, and control are beginning to let us better understand how to deploy large numbers of agents with confidence. Finally, recent work in theory and infrastructure has yielded exciting new kinds of blueprints for future systems that lie beyond the evolutionary development of current technologies. From grids to Jini (http://java.sun.com/products/jini), these approaches aim to provide a universal source of dynamically pluggable, pervasive, and dependable computing power, while guaranteeing levels of security and quality of service that will make new classes of agent applications possible.</p> <p>However, a large and ugly chasm still separates the world of formal theory and infrastructure from the world of practical nuts-and-bolts agent-system development-this is where the third research front comes in. If agent technology is ever to become as widely used as ordinary object technology is today, we must create new sorts of tools to help non-guru developers bridge the gaps between theory, plumbing, and practice. Currently, full appreciation of leading-edge developments in agent theory and frameworks requires sophisticated knowledge of speech-act theory, formal semantics, linguistic pragmatics, logic, security design, Internet and middleware technologies, distributed computing, planning, and other disciplines that are not likely to be present in a typical developer's skill set. Without good tools, rapid advances in theory and infrastructure might paradoxically attenuate rather than accelerate the adoption of agent technology as members of the developer community spin their wheels or ultimately give up in disgust. Hence we must ask: Is it possible to make development of sophisticated agents simple enough to be practical?</p> <p>Fortunately, the agents community has not completely neglected the question of tools. The DARPA CoABS program and complementary initiatives in Europe and Asia are vigorously supporting research to accelerate the development of scalable interoperable agent theory and tools, and are promoting the eventual adoption of standards through bodies such as FIPA. As part of these efforts, we are working to extend theory and create tools in two areas: agent communication and agent management. This article discusses our current research directions and preliminary results in each of these areas.</p>
J. M. Bradshaw et al., "Agents for the Masses?," in IEEE Intelligent Systems, vol. 14, no. , pp. 53-63, 1999.