Ecosystems enabling adaptive composition of intelligent services
Conference Paper
Overview
Research
View All
Overview
abstract
Intelligent services are roughly defined as pieces of
software with the capabilities of problem-solving and autonomous
composition of solutions, e.g., composition of manufacturing
processes. They are heterogeneous and distributed by design,
however in industrial settings they are constrained to local
interactions with limited room for adaptation due to the need
to lower the interoperability barrier. In this paper, we present
an approach where collections of intelligent services are treated
as ecosystems, using food chains, environments and migration
to enable adaptive compositions that generate solutions with a
higher service value chain. We present a set of experiments
demonstrating how a distributed ecosystem achieves compositions
of solutions with higher service value chains while balancing the
load and diversity of intelligent services across the ecosystem via
self-organisation. This supports the claim that implementations
of intelligent service based systems (ISBS) as ecosystems could
bring substantial benefits to industrial applications.
