Cooperation with ULB

These are the Software Languages Lab project proposals for the course PROJ-H-402.

IDEAT[1] is an integrated development environment, based on Eclipse, for the AmbientTalk[2] programming language. Both are developed in the Software Languages Lab in the VUB. At the moment, IDEAT provides primitive support for the development of mobile applications written in AmbientTalk. The goal of this project is to introduce advanced features to IDEAT by extending the current set of plugins that compose the IDE (such as: auto format, outline mode, intelligent auto-completation, integration with Android development environment, …). This may require adaptations to the AmbientTalk interpreter which is written in Java.

Contact: Carlos Noguera and Elisa Gonzalez

1. http://code.google.com/p/ambienttalk/wiki/EclipsePlugin
2. http://code.google.com/p/ambienttalk/

Current database technology focuses on stable storage of reasonably stable data. However, considering the fact that more and more items in our everyday environment are tagged with RFID tags, a new type of database is emerging. Data is dispersed onto the memory of these cheap, tiny RFID tags which can be read and written by RFID readers. When different RFID readers are interconnected in a network (e.g. a car factory where the different assembly lines are monitored using RFID) we speak of an RFID network. The difference with a normal database is that RFID tags can move in and out of range of RFID readers, rendering the data that is present in the distributed database volatile.

The idea of the proposal is to apply and adapt well-known database technology to RFID networks. Instead of manually interfacing with every single RFID reader, the idea is to develop a variant of SQL that works with the volatile data on the tags in the entire network. In sensor networks, which are similar to RFID networks, this is called network-centric programming. In this network-centric language (which is usually based on SQL), it should be possible to write queries over the entire network that keep running and of which the results are updated automatically when tags move in and out of range.

Contact: Andoni Lombide Carreto and Kevin Pinte

Model-driven engineering (MDE) is an approach to software engineering where models not only serve as documentation, but are a central artifact in the development process. Model transformation plays a central role in the manipulation of models. Currently, model transformations are used mainly to translate from a high-level model – or Platform-independent model – to a low-level model – or Platform-specific model. These transformations are executed in several refinement steps until code is reached. However, updates in the code are seldomly translated back to the high-level model. This causes the model and code to fall out of sync, and the value of the model decreases. In this project, you will work on the transformation of code changes back to the high-level model – also called Round-trip engineering – using our Instant Messenger case study.

Contact: Dennis Wagelaar

(a participatory approach to noise monitoring in real space-time)

NoiseTube (noisetube.net) is a research project which develops a new participative approach for monitoring noise pollution. The idea is to extend the current usage of mobile phones by turning them into noise sensors, thus enabling citizens to measure their own exposure in their everyday environment. Each user can also participate to the creation of a collective map of noise pollution by sharing geolocalized measurement data with the NoiseTube community. The software consists of a sound-measuring application for smartphones on the user side, and software dealing with data processing and geographical representation on the server side. The mobile application is written in J2ME and aimed at recent smartphones with Java support. Your job would be to extend the potential user set considerably by developing the application for the iPhone, which has an operating system based on Objective C rather than Java and thus requires a separate version of the NoiseTube mobile application. Successfully developed software will be used in our research project to carry out noise mapping experiments in Brussels and beyond, and in later generalisations towards air rather than noise pollution measurements.

Requirements: Java, C, interest in one of the hottest research topics currently around.

Contact: Ellie D'Hondt and Matthias Stevens