FitOptivis – From the cloud to the edge – smart IntegraTion and OPtimization Technologies for highly efficient Image and VIdeo processing Systems
The central object of the FitOptiVis project is searching for new methods and strategies for the analysis of images and video in Cyber Physical systems, which make up one of the KETs (Key Enabling Technology) of innovation.
Into FitOptivis, EOLAB has developed a reference architecture, placing the validation of the project on a water supply system in which there are: an aquifer (in which the source is accessible through an inspection well), an aqueduct (mostly underground and with some inspection wells) and a distribution tank.
In such systems, the interaction between different components is fundamental, giving particular emphasis to multisectoral applications and to the reduction of energy consumption, improving the performance.
Water supply systems are infrastructures that present security vulnerabilities, and are monitored with chemical or pressure sensors but not with cameras. This is why FitOptivis proposes an evolutionary strategy in access regulation using a control through facial recognition and analysis of the behavior of people in the vicinity of the distribution lines. The sensors and the devices used, such as drones, also allow to avoid the waste of resources and to protect the safety of operators who can avoid direct access to remote areas.
FitOptivis – realizing data analysis in real time – allows you to get better performance in the action-event mechanisms, avoiding false alarms due to weather conditions detected by the systems in vogue.
The platform is designed to facilitate the verticalization of instruments in various physical implementations, always finding a link between high performance, low power consumption and implemented features.
Scope: European project
Period of reference: june 2018 – june 2021
Scientific supervisor: prof. Luigi Raffo, professor, University of Cagliari, firstname.lastname@example.org
API: Intelligent Remote Piloting Aircraft for environmental monitoring
The API project is based on an innovative platform that employs remotely piloted aircraft (drones) in order to monitor areas at risk such as assemblies and critical areas such as aqueducts, power plants and nuclear power plants. API takes a step forward compared to the state of the art as it introduces a monitoring system that uses a fleet of mini-UAVs controlled through artificial intelligence algorithms.
The fundamental objective is the autonomous processing of relevant information on environmental safety, with an artificial intelligence system and hardware with low energy consumption.
To achieve the objective, the platform uses a series of small and low-cost drones, in communication with a base station, according to an IoT system in which distributed intelligence is the main factor of innovation.
The sensors used are of two types: optical or infrared sensors to obtain HD images and electromagnetic sensors that can provide information even in critical conditions such as fires or fog that reduce the efficiency of optical sensors. In this way it is possible to obtain a control of the dynamic mission, able to adapt to situations of different entity and degree of danger.
The drone fleet sends in real time to the ground station only the relevant information, in order to avoid the use of memory and power due to the storage of high resolution images that would not be useful. Data integration and merging systems and intelligent operational coordination systems enable the immediate intervention of an operator in emergencies.
The impact of the introduction of this technology is evidenced by the growing attention of the industry towards Smart Things in the civil sphere. The API project, developed at the headquarters of Nurjana Technologies, fits in with a view to integrating the university education to the Sardinian territorial development with important effects on employment. It is also in line with the European Data Information Access Services (DIAS) initiative, which aims to facilitate the use of satellite images.
Scope: cluster project POR FESR Sardinia 2014-2020 – Axis I – Scientific Research, Technological Development and Innovation
Scientific responsible: Pietro Andronico, Managing Director of Nurjana Technologies, email@example.com
GA-VINO: methods and technologies for innovative and sustainable management of water resources
The GA-VINO cluster project is located in the ICT and agro-industry sectors in Sardinia and is a technological platform based on a network of sensors able to monitor parameters such as the amount of water actually lost in the vineyard due to evaporation and transpiration.
The main objective is to improve environmental sustainability and reduce production costs, improving the quality of wines through analysis tools and monitoring of water resources, and consequently also reducing the hours of manual work necessary to control the water status.
In relation to the latter, in fact, the vineyard can suffer damage due to drought or excessive abundance, which involve the production of wines with different organoleptic characteristics and affect the quality and quantity of the harvest.
The technology used is characterized primarily by a simplicity of use that allows its use on a large scale and consists of:
- A hardware and software part with innovative robust wireless sensor nodes, capable of exploiting solar energy or long-life battery, which quantitatively determine the crop’s water status and intervention thresholds;
- A cloud platform for collecting and managing information useful for decision support.
Through the Internet of Things (IoT) paradigm a real connection is established between the vineyard and the cloud, and the whole technological apparatus communicates with mobile devices such as tablets and smartphones, allowing the winemaker to check the status of the vineyard in real time.
Given the multidisciplinary nature of the project, the collaboration between the University of Sassari – AGRARIA, University of Cagliari – DIEE and CNR-IBIMET- AGRIS Sardinia is a useful strategy to combine technological support with the skills and requirements of the wine sector, enhancing the peculiarities of the products typical.
The production and marketing of wines is the primary item in the turnover of Sardinian companies, both for the particular attention to quality and for the need for innovation aimed at incentivising methodologies and decision-making support in order to improve company efficiency and sustainability in the long term.
The environmental analysis of production systems is carried out through Life Cycle Assessment (LCA) studies, the most advanced and complete tool for assessing the environmental impacts associated with a product, activity or service.
Finally, another important scientific result is the assessment of the environmental sustainability of the entire company production process, which is able to highlight the possibility of undertaking a path for obtaining the company’s product certification.
Scope: cluster project
Scientific responsible: Costantino Sirca (firstname.lastname@example.org)
DoMoMEA – Tele-rehabilitation Domiciliary neuroMotor in favor of subjects with cerebral stroke with disability Moderate through Advanced Electronic devices
DoMoMEA, at the international level, is the first technological system of home telerehabilitation in favor of patients affected by stroke, who have completed the period of hospitalization. From the research project comes the possibility for the health system to reduce the difficulties due to the achievement of hospital facilities, especially in logistically disadvantaged areas.
DoMoMEA uses a sensor system for quantitative movement analysis during rehabilitative exercise, providing feedback of various types (visual, auditory and vibro-tactile). In this way the patient interacts through an interface on a TV or tablet, with exercises aimed at his disability. Therefore t is possible to customize each program according to the patient’s needs and timing.
The methodologies and resources used in the DoMoMEA project are useful elements for the development of systems in tele-rehabilitation contexts.
Role: DoMoMEA is a Top-Down Cluster project funded by Sardegna Ricerche with POR FESR funds 2014/2020 – PRIORITY AXIS I “SCIENTIFIC RESEARCH, TECHNOLOGICAL DEVELOPMENT AND INNOVATION”.
Scientific Responsible: Danilo Pani, Ph.D. – Researcher in Electronic and Informatic Bioengineering, email@example.com
ALOHA – Software framework for runtime – Adaptive and secure deep Learning On Heterogeneous Architectures
European Project (Horizon 2020)
ALOHA is the English acronym for Software framework for runtime – Adaptive and secure deep Learning On Heterogeneous Architectures. The research project’s main objective is to simplify the implementation of Deep Learning algorithms (artificial intelligence) on different low-energy computing platforms.
The main practical implications of ALOHA concern video surveillance, smart industry and bioimaging.
As part of the surveillance of structures in critical areas (for example power plants), it is possible to reduce the time needed to make the identification and tracking algorithms of objects reliable.
In the smart industry, ALOHA allows designers to optimize speech recognition for machine control, with real-time processing even in conditions of noise and reverberation.
Finally, in clinical practice, from the first results of the analysis of images obtained with computed tomography without contrast, the assistance system has recognized cases of intracranial hemorrhage with greater precision than the emergency room physicians.
The ALOHA project boasts the collaboration with 14 partners from 7 different countries and bases the analysis on the hardware during the algorithm design phase, the adaptability, the safety and the extensibility of the foreseen tools.
Role: project coordinator
Task: coordination and implementation of hardware accelerators for in-depth learning
Scientific Coordinator: Paolo Meloni, EOLAB – University of Cagliari, firstname.lastname@example.org
CERBERO – Cross-layer modEl-based fRamework for multi-oBjective dEsign of Reconfigurable systems in uncertain hybrid environments
European Project (Horizon 2020)
The CERBERO project has a range of activities that extends in various sectors regarding the creation of hardware and software modules for smart objects, particularly in assisted driving, ocean monitoring and spatial exploration.
In the case of smart electric vehicle traveling, the goal is to test and validate the driver support interface developed by the CERBERO team, interacting with the virtual guidance system developed at the Fiat Research Center (CRF). The latter, therefore, is enriched with features such as monitoring driver behavior, testing and validating a new support interface.
The CERBERO support interface integrates the simulation of road conditions, the battery status related to the use and the progress of the route in relation to the charging stations.
In this way human-machine interaction is possible, enriched by the assistance that a vehicle can provide to the driver, for example by ensuring that the battery has sufficient charge to complete the route.
Self-repairing systems for space exploration
The main objective, in this case, is to equip the systems for space exploration of self-monitoring and self-repair functionalities.
This is made possible thanks to integrated processing processes. This makes it possible to repair radiation faults or environmental conditions and downtime. Everything is at a later date in a single date of the latest data from continuous monitoring and maintenance.
Robots for monitoring the oceans on the surface and in the seabed
The Ocean Monitoring use case involves the construction of an equipment with sensors capable of producing high resolution images and videos during dives. These are “marine eyeballs” through which it is possible to obtain photographic material and live video, being connected via wi-fi to a smartphone or a PC. The marine robots are remote controlled within the range of wireless action but are also designed for self-management of navigation.
They will be useful for marine biologists to study fish fauna in places where diving is dangerous but also for monitoring the environmental conditions of the ocean floor and surfaces.
In the CERBERO project there are 11 research institutes published by 7 EU and non-European countries.
Role: project partner
Activity: provide hardware reconfiguration and rapid prototyping
Scientific coordinator: Francesca Palumbo, PhD, EOLAB – University of Cagliari, email@example.com
NEBIAS – NEurocontrolled BIdirectional Artificial upper limb and hand prosthesiS
The four-year project NEBIAS was launched in 2013 and made it possible to test a new prosthesis of the upper limb. Our laboratory has participated in the project with the realization of the circuitry to build an interface capable of making the neuro-controlled prosthesis perceived by the amputee as the natural one.
From the clinical studies conducted on selected patients, emerges a new interface that can provide a stable and very selective connection with the nervous system. From the union of different branches, mainly microelectronics and materials science, we obtain a system that, starting from the nerves with which the prosthesis interacts, detects the patient’s will to conduct hand movements in a selective way and to respond to the perceived sensation by modulating the intensity of the force.
The next objective is to miniaturize the electronic components that would not make the use of the device possible without the supervision of the research team, find an external supply system suitable for the prosthesis and adapted to the daily needs of the patients; integrate tactile feedback of sensory perceptions such as hot and cold.
Role: project partner
Task: development of implantable electronics for recording/stimulation
Scientific Coordinator: Massimo Barbaro, EOLAB – University of Cagliari, firstname.lastname@example.org
HEREiAM – An interoperable platform for self care, social networking and managing of daily activities at home
HEREiAM is a new hardware/software ICT structure that combines the ease of use of TV with the potential of broadband internet services, it allows you to monitor vital parameters, communicate with loved ones and with doctors and keep track of the agenda for taking drugs.
HEREiAM, in fact, sends reminders, connects to the smartphone of those who take care of the health status of the elderly, provides information on how to measure pressure and blood sugar and also allows you to order the expenditure directly using the remote control.
The platform is based on different hardware components:
• Android custom set-top box
• Personal smart card
• TV with HDMI
• personalized remote control
• satellite dish (useful where there is no broadband connection)
• external devices (glucometer, scale, etc.)
The set-top box, the heart of the system, of useful and integrative features such as webcam, microSD card reader, Ethernet port, Wi-Fi and Bluetooth, and to it connect, via USB, external devices. It is connected to the HDMI TV port, allowing users to access the system autonomously by inserting the smart card, avoiding procedures for filling out forms. The remote control is the only device necessary to interact with the normal operation of the TV and the platform.
The primary objective is to enable older people – people aged 60 or over – a better and more lasting self-sufficiency by providing an easy-to-use tool. According to the statistics, the potential users of this service, by 2050, will be 21.1% of the world population, which is equivalent to about two billion people.
With the increase in life expectancy, health spending also increases and the need for greater accessibility to tools to monitor their health status. HEREiAM proposes a simplification in the healthcare sector where there is no technological literacy, often due to the lack of use of virtual infrastructures or their absence.
The target was identified through demographic studies also linked to the lifestyle of three countries: Italy, Holland and Belgium. The advantages have also been identified in the size of the screen that allows its use with greater visibility, and in the social dimension given by the possibility of communicating.
Particular emphasis also on the problem of privacy: many elderly people, in fact, do not like using of video cameras for monitoring. In the case of HEREiAM, proximity sensors are used that can detect the presence of the elderly in a given environment and its mobility.
Scope: Project co-funded by the AAL Joint Program (Ref. AAL-2012-5-064) and the National Authorities and R&D programs in Italy, Belgium, Romania and The Netherlands.
Role: Project coordinator
Task: Development of the hardware and software tools operating at home, organization of the field trial in Cagliari (Italy), analysis of the results.
Scientific coordinator: Luigi Raffo, EOLAB – Università di Cagliari, email@example.com
RE.MO.TO – Movement Recovery and Telemonitoring for rheumatology patients with hand disabilities
The platform is based on sensor-assisted devices integrated into the battery-powered tele-home-care device, specifically designed for the project and through which the home patient can carry out the preconfigured exercises.
The system also provides a valid training support to the patient and transmits the data directly to the doctor, who can obtain a quantitative and qualitative report of the therapy through a software interface. The patient ID is the IMSI code of the SIM installed on the device: through this parameter, the doctor can trace back to the patient without sending personal data.
RE.MO.TO allows you to develop customized solutions in relation to the patient’s status, can be used safely both at home and in the clinic, under the supervision of the doctor. The ease of use allows the patient to be motivated in therapy, reducing cases of neglect due to lack of follow-up. At the state of the art there are no rehabilitative aids that have similar functional and non-functional characteristics.
The project also solves the difficulties of access to the NHS if the hospital facilities are difficult to reach from a logistical point of view, creating a platform that guarantees, through the internet connection, the monitoring of how the exercises are performed.
In addition to demonstrating the clinical efficacy of this implementation, RE.MO.TO offers numerous development perspectives in the field of telemedicine and rehabilitative approaches. For the hardware / software platform developed, a patent application has been filed, which in its development constitutes a new starting point for introducing different types of exercises and encouraging their application in other rehabilitation contexts.
Call for proposals: Cluster project – Fundamental or basic research projects, Regional Law n. 7 of the 7th August 2007, 2008 Annual Report
Reference period: 2010- December 2012
Responsible for EOLAB: Luigi Raffo, EOLAB – University of Cagliari, firstname.lastname@example.org
PI for EOLAB: Danilo Pani, Ph.D. – Researcher in Electronic and Informatic Bioengineering, email@example.com
ELoRA – Low-power Real-time processing of neural signals for prosthetic aids
The EloRA project presents as an innovative focus the real-time decoding of neural signals detected by the peripheral nervous system and sent in input to an active robotic prosthesis. The main application explored during the project is related to upper limb neuroprosthetics (in particular hand prostheses), in which the signal is extracted directly from the residual nerves of the amputee, through TIME (Transverse Intrafascicular Multichannel Electrode) electrodes.
Directly analyzing the modulation signal of motor activity transported by the peripheral nervous system, allows to control a high number of degrees of freedom in a more natural way compared to electromyographically controlled solutions.
At the beginning state of art, the main difficulty was to combine the need for real time, computational complexity of the algorithms and power of the embedded platforms. To achieve an efficient and effective solution, EloRA has combined bioengineering skills in signal processing with microelectronic skills, developing a system in which each processing element is specialized in performing a computational task in parallel.
Having achieved the general objective of the project, EloRA has developed a multidisciplinary approach in the analysis of biomedical signals and implementation of custom multicore architectures with low power dissipation, laying the foundations for use also in prosthetic applications different from those implemented.
Call for proposals: Fundamental research projects or basic annuity 2012 – Regional Law 7 August 2007, no. 7
Reference period: Feb 2009 – Dec 2011
PI: Danilo Pani, Ph.D. – Researcher in Electronic and Informatic Bioengineering, firstname.lastname@example.org
FetalHeart – Algorithms for non-invasive extraction of the fetal electrocardiogram
A starting point is a scenario in which there are invasive solutions – applicable only during labor – and non-invasive solutions, which do not allow to analyze the morphology of the signal as they are based on ultrasound technology.
The main problem related to the target is to be able to provide techniques that allow to perform a prenatal screeneng through the fetal ECG, aimed particularly at women who have risk factors such as congenital heart disease or familiarity with them.
The optimal gestation period goes from the 22nd week onwards, in order to allow the recording of a signal that is not affected by the small size of the fetal heart and allows timely intervention with the eventual administration of placental drugs.
The target can be extended even in the absence of such factors, as a non-invasive and inexpensive detection and screeneng tool.
Key points and objectives are identified:
- in the collection of signals obtained by applying electrodes on the pregnant women’s abdomen at different times of pregnancy, solving the problem of the lack of validation and support data;
- in the fetal electrocardiogram extraction in real time and characterized by valid quality from the clinical point of view;
- in the evaluation of the feasibility of a high performance embedded hardware system, able to realize what was proposed.
The solutions proposed by the development of algorithms able to separate the ECG signal of the fetus from the maternal one and the noise, optimizing the positioning of the electrodes, represent an important step towards the creation of innovative non-invasive diagnostic tools for fetal non-invasive monitoring childbirth.
Typology: EXPERIMENTATION PROJECTS AND TECHNOLOGICAL TRANSFER FOR BUSINESS DISTRICTS OR CLUSTERS – Distretto Sardegna ICT – POR SARDEGNA
2000-2006 MEASURE 3.13 – Research and technological development in enterprises e territory.
Reference period: Dec 2008 – Feb 2009
Scientific Responsible: Luigi Raffo, EOLAB – University of Cagliari, email@example.com
PI: Danilo Pani, Ph.D. – Researcher in Electronic and Informatic Bioengineering, firstname.lastname@example.org
MADNESS – Methods for predictAble Design of heterogeNeous Embedded System with adaptivity and reliability Support
Scope: European Project (FP7)
Abstract: The main goal of the project is to define innovative methodologies for system-level design, able to guide designers and researchers to the optimal composition of embedded MPSoC architecture, according to the requirements and the features of a given target application field. The proposed approach tackle the new challenges, related to both architecture and design methodologies, arising with the technology scaling, the system reliability and the ever-growing computational needs of modern applications.
The proposed methodologies extend the classic concept of design space exploration to:
- Improve design predictability, bridging the so called “implementation gap”, i.e. the gap between the results that can be predicted during the system-level design phase and those eventually obtained after the on-silicon implementation.
- Consider, in addition to traditional metrics (such as cost, performance and power consumption), continued availability of service, taking into account fault resilience as one of the optimization factors to be satisfied.
- Support adaptive runtime management of the architecture, considering, while tailoring the architecture, new metrics posed by novel dynamic strategies and advanced support for communication issues that have been defined.
The project proposes an extended system-level design space exploration approach. The DSE is seen as an iterative process operating on a library of heterogeneous IP cores, exposing a large number of degrees of freedom, as typical for industrial-strength components.
In order to improve the design predictability, the project will introduce a specific layer for rapid and accurate emulation, to be exploited for architectural evaluation inside the DSE process. In detail, it will allow to take into account, during system-level decision phase, the impact of the variables related with a prospective physical implementation of the architecture (e.g. wiring capacitances and delay, 2D floorplanning). Moreover, this layer will provide the capability of performing, when needed during the optimization process, a detailed estimation of the performance and the power consumption of a candidate architecture, relying on an FPGA-based environment for on-hardware prototyping. The power consumption evaluation will be obtained annotating the FPGA emulation results with energy values estimated in a “technology-aware” manner.
Role: Project coordinator
Task: Development of the FPGAbased platform for rapid emulation.
RPCT – Reconfigurable Platform Composer Tool
Scope: Project funded by regional government of Sardinia
Abstract: Modern digital technologies applied to the multimedia field have led to developing complex systems for the creation, the use and the sharing of audio/video/image pervading the consumer electronics market.
The ever increasing complexity of multimedia applications and their heterogeneity, highly characterizing the Digital Media Technologies, threaten both the standard design flow and the traditional architectural models. At the design flow level, the management of parallelism, concurrency and proper resources allocation cannot be performed anymore after the definition of the hardware substrate, otherwise causing inefficiencies in resources definition and scheduling. In this sense the RPCT project, in order to bridge the gap between the development of the hardware substrate and the mapping of applications on the available resources that negatively affects the time-to-market of this kind of systems, aims at the definition of an automatic framework ( RPCT – Reconfigurable Platform Composer Tool) for the characterization of application specific hardware systems. The RPCT framework is going to leverage on the adoption of a common formalism for both hardware and software through the use of the dataflow model of computation. A dataflow-based approach will allow software developers to model applications performance depending on the hardware characteristics, at the high-level and long before this latter is physically available, through early stages estimations exploring the cost vs. benefit of different platforms. At the same time, hardware architects using such a design flow will be able to define the optimal architecture with respect to the specific software requirements.
Role: Project coordinator
Task: Development of the reconfigurable platform