to create a software infrastructure for Ambient Intelligence (AmI) applications, that handles context at its constructive level, that works naturally with the user's tasks and activities, and that manifests the robustness and reliability that is required for future ambient systems.
While commercial IoT ecosystems exists, they are difficult to acquire and to inter-operate. On the other hand, smartphones are full of sensors and everyone has them.
The goal of this research is to create an open IoT ecosystem in which common smart devices such as phones and tablets are the main actors, being able to get information from the environment, communicate it to each other, and making it available to the users, all securely and privacy-aware. As an appropriate enabler for the subject, the system will be underpinned by software agents.
After the first year of research, the expected result is a state-of-the art multi-agent system that manages the information in a context- and privacy-aware manner, all while being flexible and lightweight. During the second year, an application will be built that uses machine learning to learn from the acquired data and detect unusual situations, such as medical emergencies.
In the Internet of Things devices are accessible via the network from anywhere in the world. An important aspect is to make data and control available the the authorized users without breaking privacy.
The goal of this research is to develop mechanisms and algorithms for data to be managed so that, while mobile through the network, is made available only to the intended / authorized users. At the same time, these mechanisms should be lightweight enough to work on resource-constrained devices.
The expected result is to have the developed mechanisms integrated into a MAS-enabled IoT ecosystem, potentially with help from the features offered by software agents.
While machine learning works great when computing resources abound, a privacy-protecting distributed implementation of context-aware, intelligent behavior requires learning on resource-constrained devices.
The goal of this research is to develop means to learn from the user's activity on smart devices only with the help of local resources.
The expected result is to be able to predict the activities of the user, using sensors on the device and also information available from local apps, in order to better assist the user in those activities and anticipate potential unwanted situations.
The goal of the project is to build a cross-platform application that is able to move from one device to another as instructed by the user. On his/her personal device, the user will slide with the finger in the direction of a PC or gesture in the direction of a PC and the application will move to the designated machine. On all machines the application will feature similar interfaces.
The project will be split into three sub-projects:
The completion of the project will get us closer to realizing the ideal deployment of AmI in which applications are independent of, and can move between devices. It will also be a good demonstrative application for the capabilities of the AmILab.