Towards a (de)composable workflow architecture to define data collection policies

Cyril Cecchinel (1, 2) , Sébastien Mosser (1, 2) , Philippe Collet (1, 2)

1 Laboratoire I3S - SPARKS - Scalable and Pervasive softwARe and Knowledge Systems
2 I3S - Laboratoire d'Informatique, Signaux, et Systèmes de Sophia Antipolis

Sensor networks are classically used in the Internet of Things to collect data, typically supporting Smart Cities or Smart Homes use cases. However, a deep knowledge of these networks is needed to properly develop applications over the deployed systems. This leads to a target mismatch: developers know how to exploit the collected data to develop large-scale " smart " systems, but do not have enough knowledge to technically enact and compose such behaviors on a given sensor network. In this paper, we envision a tooled approach that supports data collection policies management at a higher level of abstraction, fostering reuse. We discuss an architectural abstraction based on workflow concepts assisting developers in expressing data collection policies. The resulting architectures are then composable to be enacted on the same sensor network, but they can also be partially reused through a selection operator.

Keywords: Sensor network; software architecture; software reuse

Domains: Software Engineering [cs.SE]

Cyril Cecchinel, Sébastien Mosser, Philippe Collet. Towards a (de)composable workflow architecture to define data collection policies. Symposium on Applied Computing (SAC 2016), Apr 2016, Pisa, Italy. ⟨10.1145/2851613.2851962⟩. ⟨hal-01341093⟩