Before being integrated, big data must be harmonized, as it derives from heterogeneous sources, such as structured data (enterprise data, CRM, ERP), unstructured data (documents, videos, social media, sensors), and semi-structured data (EDI, XML/JSON, transactions). A data exploration process takes place to find relevant data, followed by a knowledge filtering/mining process. The methodologies used are the integration of data flows and data visualization. This research area's objectives include automatic design, guided by the information extracted from the data, and diagnosis and maintenance. A 3-step cycle (collect-analyze-optimize) will be implemented to make the rendering of big data effective.

This research area is aimed to test novel paradigms for user interaction. Specifically, new models of mixed reality, scene augmentation, and virtual reality training will be studied. The research staff will monitor all the activities with tracking and multimodal data capture. Machine learning will be applied to predict the behavior of both users and systems. 3D modeling will include computer graphics, geometry procession for digital manufacturing, and virtual/mixed reality content creation. This area aims to validate the interaction between augmented reality and virtual reality, predict gestural interaction, and monitor specific benchmarks. Novel solutions in human body modelling and shape/appearance modelling and retrieval will also be developed. The results obtained will be shared with the use of demos and scientific publications, and presentations.

This research area is implemented to develop new control and interaction methods for traditional machine tools, industrial robots and cooperative robots. Tests will measure the technical performance of both conventional and cyber-physical devices. The best sensor type will be applied to traditional robots; furthermore, different configurations and data processing algorithms will be added to identify problems, measure performance, and predict maintenance needs before faults will occur. Adequate training, applications, and technologies will be applied to suit the needs of low-cost robots. Cyber-physical systems will be equipped with new sensors and algorithms to acquire high-level decision capabilities. The objective is to fully integrate machine accuracy and human experience in the design, production, and maintenance processes.

Cloud computing brings an extra layer of vulnerability to already highly vulnerable systems. Its higher level of integration causes a risk both to the IoT paradigm and the Cyber-Physical Systems (CPSs). IoT devices are subject to attacks that exploit the lack of authentication/authorization mechanisms. Once an attack is completed, CPS undergoes physical consequences apart from IT security. This research area will test and improve formal and semi-formal methodologies to formalize, statically detect and mitigate vulnerabilities in CPSs and IoT devices and applications. Some examples of applied procedures include authentication barriers to protect IoT security and privacy, static and runtime analysis for CPS security, formal metrics to estimate the impact of cyber-physical attacks, integration of data and operation flows for web security cyber-threat intelligence with the use of statistical methods and Artificial Intelligence.

This research area will use asset identification and tracking through barcode and RFID, wireless network, and environmental sensors as starting points. The purpose is to test and develop novel systems and methodologies related to the Industrial Internet of Everything. This new concept will convey the global integration of produced objects, production machines, and environmental sensors (such as temperature, humidity, and luminosity) to be more specific. Furthermore, the manufacturing process will be formalized to track and detect errors. With the use of the Industrial Internet, errors and failures in the production chain will be identified and fixed in real-time, saving time and money.

This area is devoted to research into advanced mathematical tools to model complex processes in modern applications, like the design for self-driving vehicles, data-driven forecast, smart cities, logistics, cryptography, and biomedical applications. Other tasks will be included in the field of modelling, such as data-model comparison, analytics, and numerical simulation of models. HPC implementation and appropriate numerical schemes for control and simulation will be developed, as well as an open-source platform with downloadable material. The outcome of all research in this experimental area may include the citation of the methodologies used and the results obtained in peer-review journals, reports, communications in international conferences, and media press.