Research activities of the e-Lite group focus on designing, building and evaluating interactive and intelligent system, and testing them in realistic settings. Novel user interaction modalities, empowered by ubiquitous devices, enable the discovery of innovative methods to approach new and existing application domains. We adopt an end-to-end research approach, where we aim at exploring the whole spectrum, from devices to users, from intelligence to data representation, to develop a systemic approach.
Currently, the application area in which we are mostly active is Ambient Intelligence.
Ambient Intelligence (AmI) aims at building a digital environment where people are surrounded by intelligent and intuitive interfaces embedded in all kinds of objects. The ultimate goal of such an environment is to support people in their daily lives in a proactive, yet sensible and unobtrusive, way.
In this wide area, we mainly focus on the intelligent aspects (e.g., Machine Learning, Semantics) of the environment and on innovative interaction modalities for people, typically exploiting off-the-shelf devices and systems from the Internet of Things and Smart Home domains.
Moreover, we deserve particular attention to the Ambient Assisted Living field, in which Accessibility and Assistive Technologies meets AmI. In this domain, we experiment with novel interaction modalities for people with disabilities, with accessible games, and we build wearable and mobile systems for helping caregivers and healthcare assistants.
Accessible Games for Single-Switch Interaction
Small children with disabilities, especially in the pre-school age, need to develop their skills by using games and game-like environments. The educational effect of "serious games" enables educators and caregivers to involve the kids with fun and formative activities.
In the case of serious motor disabilities, when the interaction with a computer is severely hampered and the usage of a keyboard or pointing device is not possible, the choice of available educational games is greatly reduced.
Our research activity in this area focuses on defining new interaction modalities for accessible games in the educational and entertainment contexts, compatible also with severe motor disabilities.
The Dog Gateway was firstly developed in 2006, mainly with research purposes. It was targeted at a single home automation technology and had the goal of showing that advanced interaction with home environments was possible. This first version was named BCF, from the Italian words for "It Just Works" (Basta Che Funzioni), to further highlight its transient nature.
From 2007 to 2014, the Dog gateway has been developed and improved embedding more features, integrating new home, building and industrial automation technologies, providing a single uniform device abstraction layer based on Semantic Web technologies (the DogOnt ontology). In these years, 3 major versions have been released, the 3.0 being the latest.
Nowadays, the Dog Gateway is an open source, OSGi compliant implementation of smart environment gateway and IoT platform featuring multi-protocol capabilities, REST-based application APIs, uniform device representation, rule-based activations, real-time data elaboration (with complex event processing technologies), Xively integration (to enable social sharing of consumptions and activations), etc.
DogOnt is a novel modeling language for IDEs (Intelligent Domotic Environment), based on Semantic Web technologies. By adopting well known representations such as ontologies and by providing suitable reasoning facilities, DogOnt is able to face interoperation issues allowing to describe:
- where a domotic device is located;
- the set of capabilities of a domotic device;
- the technology-specific features needed to interface the device;
- the possible configurations that the device can assume;
- how the home environment is composed;
- what kind of architectural elements and furniture are placed inside the home.
Explore our past and completed research projects, in various topics.