wilamowski-b-m-irwin-j-d-industrial-communication-systems-2011

Industrial Agent Technology 16-11
of UAVs involved in temporally constrained mission. Ref. [12] uses a different technology, namely swarm
intelligence, for coordinating the movements of multiple UAVs. The control logic is executed autonomously
by a UAV that results in a situation where a single human is able to monitor an entire swarm of UAVs instead
of at least one human for each UAV. In Ref. [17], UAVs with small received signal strength indicator sensors
cooperatively work together to locate targets emitting radio frequency signals in a large area. Ref. [11]
describes a decision making partnership between a human operator and an intelligent UAV. Here, the intelligence
is provided by a MAS. The MAS controls the UAV and self-organizes to achieve the tasks set by the
operator, however, also through interaction via what they call a variable autonomy interface with the operator.
In all these projects, especially the autonomy and intelligence aspects are realized by agent technology.
Autonomy, however, requires a lot of communication in order for a set of autonomous units to achieve their
goals cooperatively. Thus, communication and decision-making plays a major role in those projects.
Related to supply chain management is logistics in general. In logistics, the information and data
required for efficient planning are, for various reasons, typically not available centrally. Quite a number
of systems have been deployed in this area from which only two will be presented briefly. The Living
Systems/Adaptive Transport Network (LS/ATN, cf. [40]), provided by Whitestein Technologies, for
example, to ABX, a European logistics company, is an agent-based system for network planning and
transportation optimization for charter business logistics. Based on a simulation study that was performed
by Whitestein on the basis of 3500 transportation requests, 11.7% cost savings were achieved. The other
very successful commercial deployment is the Magenta i-scheduler, an intelligent, event-driven logistics
scheduling system based on Magenta agent technology. It is characterized by a number of unique,
advanced features such as real-time, incremental, continuous scheduling and schedule improvement,
scalability from small to very large enterprise networks, which also includes balancing of their conflicting
interests, multi-criteria schedule analysis, and rich decision support for the user. The system is
deployed already in a number of industrial applications. As a scheduling/logistics system for Tankers
International, it provides intelligent support in the scheduling of a 46-strong very large crude carrier
(VLCC) fleet. I-scheduler implements virtual marketplaces with about 1000 agents running during a
typical execution of the system (cf. [19]). I-schedulers were also deployed in several road transportation
applications with several U.K. road logistics operators (cf. [21]). The application reported in [20] copes
with about 1500 orders daily, in 650 locations, 150 own trucks, and, additionally, 25 third-party carriers.
I-Scheduler assigns an autonomous agent to every player in the transportation system and, additionally,
tasking agents to obtain the best possible deals for their clients. Players include the transportation
enterprise as a whole and all individual transportation demands and resources, like drivers and resource
owners. Demands and resources that have common interests self-organize into groups represented by
a single agent. Agents may decide to compete or cooperate depending on prevailing circumstances. To
construct a schedule, a formal description of business domain knowledge (Ontology), of particular
situations (a Scene), of agent goals (e.g., to increase profit), and of real-world events are used.
Another prominent area for agent-based systems is virtual organizations (VO) or enterprises (VEs) (see, e.g.,
[14] for a nice overview). VOs may deeply integrate all manufacturing and supply chain aspects, sales networks,
as well as suppliers, customers, and third-party maintenance and coordination. Here, again, the cooperation
and coordination aspect of such conglomerates stays in the center of what agents are supposed to provide.
More recently, research tried to add more semantics to industrial application systems. This usually
means that ontology-based concepts are integrated. A nice overview about this trend is given in [22].
An example for a prominent and successful approach is presented in [23]. The OOONEIDA consortium
provides a technological infrastructure for a new, open knowledge economy for automation components
and automated industrial products. Their framework aims to provide interoperability on the
hardware as well as on the software at all levels of the automation components market, that is, from
device and machine vendors to system integrators up to the industrial enterprises. In the center of their
approach are the so-called searchable repositories of automation objects. Here, each player can deposit
its encapsulated intellectual property along with appropriate semantic information to facilitate searching
by intelligent repository agents.
© 2011 by Taylor and Francis Group, LLC