Research 2001 - 2005
The term Ambient Intelligence (AmI) has been defined by
the ISTAG as ’the convergence of three major key technologies: ubiquitous
computing, ubiquitous communication, and interfaces adapting to the user’.
Ubiquitous Computing is about the integration of microprocessors in everyday
objects (ordinary appliances, furniture, clothes, toys and paints). Computers
are made available by means of the physical environment, but in an invisible
way for each person that is present there. ’Smart’ objects,
that is integrating embedded technology, must be enabled to communicate
with each-other by means of ad-hoc and wireless networking, thus the need
for Ubiquitous Communication technology, connecting together heterogeneous
networks. Technology is nothing if contents cannot be accessed and in
a userfriendly manner. Thus it is urgent the call for Intelligent User
Interfaces that enable the inhabitants of the AmI environment to control
and interact with the environment in a natural (gaze, voice, gesture,
touch) and personalized (preferences, context, profiling) way.
My research focuses on this third key technology. Scope of this research
within the realm of AmI is exploring Interfaces for AmI applications.
We focus on two independent guidelines for design and implementation of
IUI: wearability and usability. To experience advantages coming from personal
user devices, their wireless and computational capabilities, we implemented
an interaction system for Immersive Virtual Reality Environments reconstructing
Cultural Heritage sites. The interaction system is based on usage of PDAs
and has been extended to mobile phones. The mobile device enables navigation
and object manipulation in real-time rendered Virtual Environment displayed
on a (semi)-cylindric screen. At the same time it is a mean of interaction
with other people present in the room. The main advantage in this innovative
mix of technologies is that portable devices are easily handled and enable
a friendly interaction to the majority of users, with a reduction in learning
time needed for accessing the contents. Moreover in a Virtual Heritage
context, portable devices offer the opportunity to visualize, in a non-invasive
way, sources i.e. photos, drawings, plans, texts, etc.- necessary to validate
virtual reconstructions and to explain to the user the real cultural value
of the virtual world (s)he is visiting. We also explored natural interface
domain in the particular research field of gestures and movements recognition.
As in Virtual Reality, also in Smart Environment, the interaction methods
used affect the usefulness of being surrounded by a responsive virtual/physical
world thanks to embedded technology. Interfaces contribute to the feeling
of immersion and effectiveness. Our work addresses the problem of finding
a low-cost/low-power solution based on commodity components to enable
a simple but effective gestural interfaces. We first explored inertial
sensors and their usage for capturing tilt and movements. We also investigated
the feasibility of a biometric interface based on gestures. Then, we designed
and implemented a wireless sensor network to be worn by the user, leading
to a Body Sensor Network. Because of their small size, in fact, the wireless
communication nodes implemented can be attached to the user to create
a custom input device or for monitoring purposes.
1. Personal Devices as Interaction Platform:
Enabling users interaction with a smart environment through PAN
Case study: PDA and Immersive Virtual Heritage
2. Natural interaction: using tilt
to capture human movements and for gesture recognition.
Case study: MOCA - A Low-Power Motion Capture System with Integrated Accelerometers
3. Using gestures as biometric: a feasibility study
4. WiMoCa: a Wireless Sensor Node for Body Area Networks
5. Energy-Efficient Protocols for Wireless Sensor Networks
6. Research cooperation with HP Italia: project of an Intelligent Museum
7. Bio-feedback Wireless Wearable System
8. LAICA (Web site)
9. SUMMIT (Web site)
Personal Devices as
Interaction Platform: Enabling users interaction with a smart environment
through PAN (top)
Case study: PDA and Immersive Virtual Heritage
Virtual Heritage (VH) is a promising and challenging application domain for advanced digital communication and processing. The rapid evolution of hardware and software pervasive computing technologies creates many exciting opportunities for bringing VH applications to a new level of user participation, where highly interactive, multimedia data streams can be exchanged, in real-time, among untethered users working together in a complex virtual environment. This work demonstrates how the integration of several advanced technologies (mobile hardware platforms, high -performance rendering engines, platform-independent software development toolkits) enables the development of innovative VH applications, where peer-to-peer multimedia interaction takes the center stage.
Natural interaction:
using tilt to capture human movements and for gesture recognition. (top)
Case study: MOCA - A Low-Power Motion Capture System with Integrated Accelerometers
Motion capture is an emerging technology enabling the design
of natural user interfaces for wearable devices based on gestural recognition.
However, costs and energy requirements
are critical factors to enable their diffusion to low-end wearable systems.
Current commercial products do not match these requirements. For this
reason, we developed a low-cost/lowpower wearable motion tracking system
based on integrated accelerometers called MOCA (Motion Capture with Accelerometers).
Our system is composed by sensing units connected to a control/acquisition
board responsible for reading and preprocessing data and a mobile terminal
running the recognition
algorithm. Experiments performed to validate accuracy, power consumption
and real-time performance demonstrate low-power and flexibility features
of the proposed tracking system as well as its effectiveness as input
interface. We used the system for navigation in virtual spacesand in a
3D game application as a test-bed to evaluate the effectiveness of the
interface.
Using gestures as biometric: a feasibility study (top)
The present work is about the investigation done to test the feasibility of a person verification system using gestures as biometric. Gestures are captured through inertial sensors. If we can distinguish one person from another by the way she/he gestures, technology embedded in a Smart Environment (household, workplace, a public space) can provide contextual and adapted reaction to the presence of a certain user, transparently from her/him. The technology used to capture the gesture "signature" can be the same exploited for other tasks. For example sensors distributed on the user body can provide data to recognize the user without her/his explicit intervention and at the same time can act as a input pointing device based on explicit user gesture commands. In the feasibility study we are leading, data analyzed for the purpose are accelerations in three orthogonal directions, coming from a prototype box acting as a palmtop computer. A sample of gestures is processed through supervised and unsupervised dimensionality reduction techniques, leading to encouraging results.
WiMoCa: a Wireless Sensor
Node for Body Area Networks (top)
This work is about the design and implementation of a wireless
sensor node for a Motion Capture system with Accelerometers (WiMoCA).
It is composed by a tri-axial integrated
accelerometer, a microcontroller and a wireless transceiver. The use of
a single integrated tri-axial accelerometer allows to overcome inaccuracies
of orthogonally mounted biaxial accelerometers and to build a smaller
device compared to other sensors presented in literature. WiMoCA nodes
have been exploited to build a Wireless Body Area Sensor Network (WBASN),
that allows to implement a wireless/ wearable distributed gesture recognition
system where nodes are mounted on many parts of the human body. We describe
the hardware architecture and all the software layers supporting the recognition
system. We also show characterization experiments on WiMoCA nodes that
highlight how their performance and power consumption levels make them
suitable to HCI applications.
Energy-Efficient Protocols for Wireless Sensor Networks (top)
We implemented two MAC protocols for our WiMoCA
system, that have complementary features. We are also exploring standard
communication protocols such as Zigbee. The first one is a collision free
protocol suitable to constant rate traffic conditions, where nodes are
polled very often. Moreover, this protocol is suitable for networks with
a constant number of nodes, as the one we target in this work. The second
is a collision avoidance protocol which is more suitable to event driven
applications and more
dynamic networks, where the number of nodes is variable.
Another relevant difference is that the first one uses a
cross-layer application driven approach, while the second
one implements a traditional scheme where applications are
not aware of medium access control. The last one has been
implemented to be compliant with a fully layered approach
and is suitable to be eventually coupled with a transport
layer in future network evolutions.
Research cooperation with HP Italia: project of an Intelligent Museum (top)
We envision a scenario in which a user can move around museum rooms with a personal device (cellular phone or PDA) and download information related to paintings, sculptures and other pieces of art close to the user (location awareness). The user will have different ways to handle information concerning the exposed objects. For example, a vast amount of text, images, multimedia material (audio and video) related to the collections may be available from the museum information gateway. Instead of just supporting user controlled navigation in the available information (e.g., in hypertext format) we will enable a natural and interactive user experience such as 3D virtual object manipulation and virtual walking through ancient sites (natural HCI). Moreover we will allow dynamic and flexible association of user sessions to different input and output devices (for instance, different visualization windows will be available to the user, depending on her location in the museum).
Bio-feedback Wireless Wearable System (top)
Nowadays, collecting information from the body
is becoming crucial in several domains including medicine, natural interaction
for home-wide automation and communication, virtual reality applications,
sports, and
ergonomics. The lack of suitable solutions for fusing information deriving
from different parts of the body limits the understanding of the interaction
among different physiological systems and the possibility to actively
improve
physical performances and the interaction with the environment.
The aim of the WI-BODY project is to test and release a next-generation,
smart, portable system which enables:
(1) synchronous monitoring of multiple biomedical information and (2)
active control of physical performance
using principles such as biofeedback and functional electrical stimulation.
This system is intended to provide a
valid, economically advantageous alternative to state-of-the-art devices
(when already existing), or an original,
powerful tool in the field of body function monitoring and control.
What we would like to deliver in the end of the WI-BODY project is a prototype
system that acts as an einclusive
interface with the environment and the society, where health and mobility
are intended as primary keys to
ensure equal access, independent living, and participation for all in
the Information Society. This project will test
the innovative WI-BODY system architecture in different fields including,
but not limited to, preventive and
rehabilitation medicine, sport training, remote consulting, and domotics.
WI-BODY system architecture makes
systems unlike they are now: easy-to-wear, easy-to-use, active anywhere,
anytime.
The goal of the project will be achieved by developing a general-purpose,
distributed, modular architecture of
network cells able to communicate with each other and equipped with a
portable logic. Network cells will
incorporate biosensors or bioactuators, either mounted on body surface
or implanted.
Long term work will be needed to design and get together the different
components involved (network cells;
communication protocols; embedded real-time algorithms for actuator control;
signal processing algorithms; data
warehousing, web-based data access) and in-vivo accurate validation will
be required before the proposed solution
will be acceptable for industrial or commercial take-up.
Ultimately, this project could have important outcomes on the quality
of life of European citizens. These new
system will empower persons with disabilities and aging citizens to play
a major role in society and will help
increasing their autonomy and realizing their potential.
LAICA (Web site) (top)
Il progetto L.A.I.C.A. (Laboratorio
di Ambient Intelligence per una Città Amica) si propone di svolgere
attività di ricerca nei settori delle tecnologie dell’Ingegneria
dell’Informazione e della Scienza della Comunicazione per creare
soluzioni di Ambient Intelligence a livello urbano. E’ un
progetto della durata di 2 anni, finanziato dalla Regione Emilia Romagna
nell’ambito dell’iniziativa 1.1 del Piano Telematico Regionale
– Programma operativo 2002, che vede coinvolti il Comune di Reggio
Emilia, responsabile delle attività del progetto, alcuni dipartimenti
dell’Università di Modena e Reggio, dell’Università
di Parma, dell’Università di Bologna, WTI, Regulus, AGACTEL,
OT Consulting, Sirti. Il team di ricercatori accademici da una parte (UniMORE
DII, UniMORE DISMI, UniMORE DSC, UniPR DII, UniBO DEIS), le disponibilità,
gli obiettivi e le esigenze della municipalità reggiana dall’altra,
sono catalizzatori di questo progetto che non possono però esimere
dalla compresenza di forti interessi industriali. Per questo nel progetto
sono confluiti, fin dall’inizio della STORIA di L.A.I.C.A., diverse
realtà aziendali e partner industriali radicati sul territorio
(WTI, Regulus, AGACTEL, OT Consulting, Sirti) che già hanno competenze
di rilievo nell’ambito della realizzazione di sensori visuali, di
sistemi software per servizi e pubblica amministrazione, di gestione di
rete e del supporto alla comunicazione.
Le attività di ricerca e di sviluppo connesse al
progetto L.A.I.C.A. possono essere descritte ipotizzando di avere a disposizione
nella città di Reggio Emilia una rete di grande dimensione di elementi
sensoriali intelligenti, ossia dotati di capacità di elaborazione
dati, oltre che di acquisizione e trasmissione. Tali sensori (principalmente
video) sono collegati in rete tra loro attraverso cablaggio fisico (MAN)
o reti wireless e sono in grado di comunicareinformazioni acquisite e
di essere interrogati mediante software distribuito e dinamico.
Il sistema di Ambient Intelligence ha, quindi, la capacità di essere
interrogato puntualmente dall’utente privato o dalla Pubblica Amministrazione
che, attraverso un terminale mobile di ultima generazione o attraverso
una interfaccia Internet, può collegarsi in modo diretto e richiedere
i dati elaborati: esempi possibili sono le immagini e i video provenienti
da sensori visuali, lo stato del traffico ad un incrocio, lo stato dei
servizi pubblici o dei parchi cittadini.
A livello globale l’intero sistema è capace di interagire
e di raccogliere tutte le intelligenze distribuite sull’ambiente
urbano (tramite un middleware), che vengono poi sintetizzate per definire
forme di comunicazione globali per tutta la città. In questo ambito,
il progetto prevede la realizzazione dell’infrastruttura hardware
e software e di un centro di gestione centralizzato delle informazioni
sullo stato globale della città in grado di erogare servizi sia
agli uffici della pubblica amministrazione che direttamente al cittadino
mediante diffusione broadcasting. Il progetto verte in maniera primaria
sulla ricerca industriale (visione artificiale, sensori e middleware).
Parziale eccezione sono le attività di sperimentazione nel corso
delle quali i prototipi (software ed hardware) saranno installati ed utilizzati
sul territorio comunale, alcuni dei quali richiedendo quindi una quantità
di ricerca pre-competitiva tale da garantirne robustezza ed affidabilità.
SUMMIT(Web site) (top)
Il progetto SUMMIT (Servizi Ubiquitari MultiMediali per l'Innovazione tecnologica e Telematica) è un progetto di ricerca industriale che prevede la realizzazione di attività di ricerca in ambito multimediale. SUMMIT è cofinanziato dalla Regione Emilia-Romagna (Delibera della Giunta Regionale n. 1598 del 30 luglio 2004) nell'ambito della Misura 1.1 del Piano Telematico Regionale. L'obiettivo che si intende perseguire è quello di sviluppare una piattaforma di tecnologie abilitanti, comprendenti software, hardware e reti, in grado di permettere l'implementazione, in tempi brevi e a costi competitivi, di applicazioni multimediali in campi molto diversi come la sanità, il commercio, il mondo della ricerca e la Pubblica Amministrazione. Coerente e sinergico con le strategia attuata dalla Regione Emilia-Romagna per la valorizzazione delle risorse in campo multimediale presenti nel territorio, SUMMIT si vuole costituire come il primo nucleo di una delle componenti del Centro per un Distretto della Multimedialità, ovvero quella denominata MultiMediaLab, dedicata a ricerca e sviluppo, che vuole essere anche interfaccia verso gli utenti e le imprese. Le applicazioni previste, inoltre, potrebbero supportare la nuova rete di laboratori del PRRIITT (Programma Regionale per la Ricerca Industriale, l'Innovazione e il Trasferimento Tecnologico).