Projects using directly or indirectly the SmartCampus infrastructure

Captah

Ivan Logre, Sébastien Mosser, Michel Riveill

Ivan Logre is focusing his PhD researches on a multi-role approach to bridge the gap between requirements, sensor data and visualization dashboard.

Using a composition-based approach to support domain variability, leading to a DSL of visualization composition, it allows one to design a dashboard by reduction of the visualization domain capabilities.

Ongoing work includes design of languages dedicated to visualization requirements (at the domain expert level of abstraction) and data polishing (virtual sensor, sampling, offset, …)

More information on Ivan Logre website

Composition of Data collection policies on a sensor network

Cyril Cecchinel, Sébastien Mosser, Philippe Collet

Facing the problem of managing tremendous amounts of data, a commonly approach is to rely on sensor pooling and to push data collected by sensor in a cloud-based platforms. Consequently, sensors cannot be exploited at the same time and one needs to rely on data mining solutions to extract and exploit relevant data according to usage scenarios.

To develop software that fully exploits an large scale sensing infrastructure, it must be considered as a white box. But the developer tasks is then more complex as they have to deal with tedious low-level details of implementation out of their main business concerns.

This project proposes a tooled approch named COSmIC (Composition Operators for Sensing InfrastruCtures) that tackle issues encountered by developers when they are using small-scale (e.g. University campus) sensor networks (lack of reusability, no sensor sharing, platform heterogenity and low-level programing)

Link to the COSmIC tool

Sensor Simulation Framework

Romain Alexandre, Cécile Camillieri, Fabien Foerster, Jérôme Rancati, Sébastien Mosser

Cloud-based middleware should be designed to receive data from an increasing number of sensors. Most of the time, developpers do not have the budget to buy hundreds of sensors and bulding sensor networks is not a part of their business concerns. Therefore, testing the scalability of their products can become an issue.

This project brings to developpers a framework allowing the simulation of sensors. Two kinds of simulations are provided:

  • The replay of previously captured sensor data
  • Predict the data that could be provided by sensors, using the definition and use of mathematical laws

These simulations can be performed in real time (e.g. to test the retrieval function of a middleware) or in virtual time. This latter allows to generate several months of data in a fast time in order to fill the middleware’s database.

Link to the simulator framework

Sensor Visu Like

Christian Brel, Emmanuel Pouthier, Anne-Marie Pinna-Dery, Philippe Renevier Gonin