Date: February 5, 2020
Time: 10:00 AM - 11:00 AM
Location: CSE 404
Host: Prabhat Mishra
Virtual Coaching for Empowering Pre-Frail Elderly in Daily-Life Activities
Speakers: Professor Nicola Bombieri and Professor Graziano Pravadelli – University of Verona, Italy
Abstract: Assisted living technologies based on embedded devices have demonstrated to be effective in alleviating the social and health-care issues related to the continuous growing of the average age of the population. Many smart applications, devices and systems have been developed to monitor the health status of elderly, substitute them in the accomplishment of activities of the daily life (especially in presence of some impairment or disability), alert their caregivers in case of necessity and help them in recognizing risky situations. However, less effort has been spent in designing smart solutions for empowering and supporting the self-efficacy of elderly. Thus, it is important to define low-cost, non intrusive, and ubiquitous virtual coaching systems to support elderly in the acquisition of new behaviors (e.g., taking pills, drinking water, finding the right key) necessary to cope with needs derived from a change in their health status and a degradation of their cognitive capabilities as they age. During the first part of the talk, we present a coaching system that 1) exploits BLE smart tags to transform common objects of the daily life used by elderly (e.g., pill box, bottle of water, keys) into smart objects, 2) monitors their usage according to the elderly needs, and 3) incrementally guides them in the acquisition of new behaviors related to the use of such objects. Also, cameras can be used as an alternative to smart objects for monitoring elderly activities. In this case, the main challenges consist of preserving data privacy while guaranteeing real-time and energy efficient data processing. The second part addresses the main issues related to the porting of such computer vision applications on low-cost embedded boards for “at-the-edge” data processing. The talk shows how we modified and customized a typical computer vision application (ORB-SLAM) combined to an inference application for object detection (GoogleNet CNN) to satisfy non-functional constraints in a widespread off-the-shelf embedded board (NVIDIA Jetson TX2).
Graziano Pravadelli, Ph.D. in computer science, IEEE senior member, IFIP 10.5WG member, is full professor of information processing systems at the Computer Science Department of the University of Verona (Italy) since 2018. In 2007 he co-founded EDALab s.r.l. (Italy), an Italian SME whose mission consists of giving support for the innovation and technology transfer in embedded system modeling and verification. His main interests focus on modeling, simulation and semi-formal verification of HW/SW embedded systems and cyber physical systems, with particular regards to virtual prototyping, correct-by-construction embedded SW generation, TLM and RTL modeling, abstraction and refinement techniques, mutation analysis and mutation testing, functional qualification and assertion-based verification. More recently he started to study the application of embedded systems to develop virtual coaching platforms for people with special needs.
In the previous contexts, he collaborated in several national and European projects and he published more than 100 papers in international conferences and journals. He is involved in the steering committee of VLSI-SOC and FDL conferences, and he served as member of the program committee of many other events, including DATE, DAC, and CODES+ISSS. Since 2015 he is an expert evaluator for ANVUR (Agenzia Nazionale di Valutazione del Sistema Universitario e della Ricerca), the Italian agency for the evaluation of the academic and research system. Since July 2017 he is President of the Quality Assurance Board (Presidio della Qualità) of the University of Verona.
Nicola Bombieri is currently an associate professor with the Department of Computer Science, University of Verona, Italy. His research interests include parallel and heterogeneous architectures, parallel computing, and parallel programming languages. He develops techniques to efficiently parallelize software applications for multicore, manycore, heterogeneous architectures targeting performance, power, and energy efficiency. His research field also includes electronic design automation (EDA) applied to Smart Systems modeling and verification, automatic generation of embedded parallel software, hardware description languages (HDLs), EDA applied to Systems Biology for network modeling and simulation. He serves as a technical program committee member, program chair, workshops/special sessions chair at ACM/IEEE conferences like DAC, DATE, ICCD, MCSoC, SIES, ECSI FDL, CODES/ISSS, MEMOCODE, DSD, VLSI-SoC, ETS. He is author of more than 100 publications in international journals and conferences. He is an editor of two books. He founded and is head of the PARCO Lab at the Department of Computer Science, University of Verona. The PARCO Lab, which goal is the research and development of advanced parallel programming techniques for parallel architectures has been awarded by NVIDIA Corporation and, currently, it hosts 5 PhD students, 1 PostDoc, 3 Master students, and 8 intern students (for bachelor and master degree stage). The PARCO Lab serves as multidisciplinary research laboratory for applying advanced and parallel architectures to Embedded Systems, Robotics, Artificial Intelligence, Bioinformatics, Systems Biology, and Molecular Biology with national and international research groups.