Meet our projects
Explainable Manufacturing Artificial Intelligence
Horizon 2020 / ICT-38-2020
XMANAI aims at placing the indisputable power of Explainable AI at the service of manufacturing and human progress, carving out a “human-centric”, trustful approach that is respectful of European values and principles, and adopting the mentality that “our AI is only as good as we are”. XMANAI, demonstrated in 4 real-life manufacturing cases, will help the manufacturing value chain to shift towards the amplifying AI era by coupling (hybrid and graph) AI "glass box" models that are explainable to a "human-in-the-loop" and produce value-based explanations, with complex AI assets (data and models) management-sharing-security technologies to multiply the latent data value in a trusted manner, and targeted manufacturing apps to solve concrete manufacturing problems with high impact.
Search & Rescue
Emerging technologies for the Early location of Entrapped victims under Collapsed Structures & Advanced Wearables for risk assessment and First Responders Safety in SAR operations
Horizon 2020 / SU-DRS02-2019
Apart from earthquakes that usually result to catastrophic structural collapses, with many people entrapped or killed (e.g. Indonesia 2018, Japan 2011, Haiti 2010, Italy 2009, Greece 1999), there are also other causes that may result into a building’s collapse, such as an accidental explosion or a terrorist attack (e.g. 9/11) in public areas or critical infrastructures (airports etc.). Moreover, natural disasters like earthquakes may trigger technological disasters, such as industrial chemical release or even fires; this dynamic or “domino effect”, as it is called may pose tremendous risks to the countries and communities and hence it is a great challenge to cope with by the first responders and relevant organizations of civil protection. First responders and rescuers need specialized instrumentations, available to all times, easily accessible that meet stringent requirements in terms of detection accuracy, quick localization, and reduction of false alarms. The S&R project will design, implement and test through a series of large scale pilot scenarios a highly interoperable, modular open architecture platform for first responders’ capitalising on expertise and technological infrastructure from both COncORDE and IMPRESS FP7 projects. The governance model of S&R will be designed to operate more effectively and its architectural structure will allow to easily incorporate next generation R&D and COTS solutions which will be possibly adopted in the future disaster management systems. The Model will also support a unified vision of the EU role and will provide a common framework to assess needs and integrate responses. The framework will enable supportive approach using a wider range of decisional support features and monitoring systems and will also give to first responders an effective and unified vision of (a) the dynamic changes going on during event’s lifetime and (b) the capabilities and resources currently deployed in the field.
More info at:
A Universal Cyber Security Toolkit for Health-Care Industry
Horizon 2020 / SU-TDS-02-2018
SPHINX aims to introduce a health tailored Universal Cyber Security Toolkit, thus enhancing the cyber protection of the Health and Care IT Ecosystem and ensuring patients’ data privacy and integrity
Hospitals and Care Centres are prime targets of cyber criminals, especially concerning data theft, ransomware, man-in-the-middle and phishing. This reflects the need of Healthcare Institutions for a Holistic Cyber Security vulnerability assessment toolkit, that will be able to proactively assess and mitigate cyber-security threats known or unknown, imposed by devices and services within a corporate ecosystem.
SPHINX aims to introduce a Universal Cyber Security Toolkit, thus enhancing the cyber protection of Health IT Ecosystem and ensuring the patient data privacy and integrity. It will also provide an automated zero-touch device and service verification toolkit that will be easily adapted or embedded on existing, medical, clinical or health available infrastructures.
The SPHINX Toolkit will be validated through pan-European demonstrations in three different scenarios at different countries (Romania, Portugal and Greece). Hospitals, care centres and device manufacturers participating in the project’s pilots will deploy and evaluate the solution at business as usual and emergency situations across various use case scenarios.
A novel holistic approach for hardware trojan detection powered by deep learning
Horizon 2020 - Grant agreement No 777222
Hardware trojans impact our everyday life and may even cause life threatening situations. Unlike other errors and malfunctions, trojans are inserted deliberately. Apart from insider attacks, the economically driven outsourcing of production steps to third party contractors enlarges the attack surface dramatically. Today, there is no single tool that can provide a holistic approach for hardware trojan detection covering both pre-silicon verification and post-silicon testing. HERO takes a significant step forward by developing an artificial intelligence empowered approach that will be capable of identifying vulnerable regions in IC designs as well as perform post-silicon validation to cover trojans of different types and sizes under large parameter variations. The deployment of HERO technology is expected to bring a breakthrough in numerous IC-related sectors including cell phones, digital cameras, microelectromechanical systems, photonics and bioelectronic devices.
in collaboration with TUM
A Fully Automated Identification of Skin Morphology in Raster-Scan Optoacoustic Mesoscopy
Identification of morphological characteristics of skin lesions is of vital importance in diagnosing diseases with dermatological manifestations. This task is often performed manually or in an automated way based on intensity level. Recently, ultra-broadband raster-scan optoacoustic mesoscopy (UWB-RSOM) was developed to offer unique cross-sectional optical imaging of the skin. In collaboration with TUM and Prof. Ntziachristos, AIDEAS is working on a machine learning approach to enable, for the first time, automated identification of skin layers in UWB-RSOM data. The proposed method, termed SkinSeg, has been applied to coronal UWB-RSOM images obtained from 12 human participants. SkinSeg is a multi-step methodology that integrates data processing and transformation, feature extraction, feature selection and classification. Various image features and learning models were tested for their suitability at discriminating skin layers including traditional machine learning along with more advanced deep learning algorithms. An SVM-based post-processing approach was finally applied to further improve the classification outputs. Random forest and convolutional neural networks provided promising results at a level of approximately 85%. The application of the proposed post-processing technique was proved to be effective in terms of both testing accuracy and 3D visualization of classification maps. SkinSeg demonstrated unique potential in identifying skin layers. The proposed method may facilitate clinical evaluation, monitoring and diagnosis of diseases linked to skin inflammation,
Application of machine intelligence for osteoarthritis diagnosis and prediction
Osteoarthritis is the most common form of arthritis in the knee that comes with a variation in symptoms’ intensity, frequency and pattern. Knee OA (KOA) is often diagnosed using invasive and expensive methods that can measure changes in joint morphology and function. Early and accurate identification of significant risk factors in clinical data is of vital importance in diagnosing KOA. AIDEAS is working on a machine intelligence approach to enable automated, non-invasive identification of risk factors from self-reported clinical data about joint symptoms, disability, function and general health. Different machine learning and deep learning algorithms have been tested and compared in terms of multiple criteria e.g. accuracy, per class accuracy and execution time. The results suggest that machine intelligence and especially deep learning may facilitate clinical evaluation, monitoring and even prediction of knee osteoarthritis.