Technological advances in miniaturized, ultra-low power embedded systems, in communication protocols and in data mining techniques are leading to disruptive innovations. Smart cities, eHealth or Industry 4.0 - Internet of Things (IoT) environments have already started to deeply modify established industries. Nowadays, every computer engineer and researcher should seriously think about how to embrace the opportunities and challenges offered by smart connected technologies. This course will cover three main themes: Ultra-Low Power (ULP) computing platforms for IoT (hardware designs, power/energy requirements), Communication for IoT (protocols, networks, latest standards) and applications-oriented optimizations of IoT systems (examples of wearable and smart home IoT case studies with data analysis and embedded machine learning). The first objective of the course is getting a comprehensive overview of IoT terminology, concepts and latest trends in the development of ULP wearable systems including platform-based designs and power-aware system optimizations. The second objective is to understand the main challenges and trade-offs in ULP IoT system designs related to edge computing vs. data communication to the cloud infrastructure in IoT networks (protocols, standards, wireless transmission, etc) in different examples and case studies (in medical, wellness, smart homes or Industry 4.0 applications).
According to the main goals described above, the course is structured in three different modules: ULP IoT platforms and power management, communication, and application-oriented IoT designs.
The first part of the course will be dedicated to "ULP computing platforms for IoT" for ultra-low power wearable systems, with sub?topics ranging from design principles and approaches, to the discussion about available platforms and development tools in the IoT space. Different examples of hands?on demos to illustrate the different features of the existing platforms will be presented during this first module of the course. The participants will get familiar with the main terms and concepts (active vs leakage power, output quality, computing error rates, classification quality, lifetime, availability rate, etc.) that are relevant for IoT platforms and will be used during the following modules of the course. Then, this module will cover how to design complete ULP wearable IoT platforms (exploring single- vs multi-core based designs, memory and synchronization choices) that can operate with minimal energy according to the target application. Also, it will be covered the set of state-of-the-art power management and energy harvesting techniques for ULP wearable IoT architectures. The demonstrations of this module are focused on examples of different energy-efficient IoT wearable and home automation systems (Shimmer, Apple Watch vs other smartwatches in the market, TI Sensor Tag, etc.)
The main topic of the second module is entitled "communication for IoT": lectures will cover the main issues and challenges related to new ULP protocols, management and optimization of communication for wearable wireless systems and IoT networks. This module will describe the essential concepts and transmission schemes behind current standards and introduce the basics of future emerging communication technologies and signaling schemes relevant to ULP IoT systems. The module will be complemented with an analysis (regarding power consumption and security) of the interaction of IoT devices and cloud services. The demonstrations of this module will be focused on the design trade?offs between different protocol standards (like ZigBee or Bluetooth LE, and LoRA or SigFox, etc.), as well as secure IoT sensors communication with cloud services by using as case study the Amazon Web Services (AWS) cloud infrastructure.
The third module of the course is application-oriented optimization of ULP wearable and smart home IoT systems. It includes dedicated examples and theory on how to build the ULP software for on-board signal processing, feature extraction and hierarchical machine learning classification approaches for energy-scalable ULP IoT platform design, with special focus on wearable IoT systems. The participants will have the opportunity to learn the state of the art and advances on approximate-computing in the context of ULP wearable IoT systems for different applications and case studies in the fields of wearables and smart home appliances.
David Atienza is associate professor of EE and director of the Embedded Systems Laboratory (ESL) at EPFL, Switzerland. He received his MSc and PhD degrees in computer science and engineering from UCM, Spain, and IMEC, Belgium, in 2001 and 2005, respectively. His research interests focus on system-level design methodologies for high-performance multi-processor Systems-on-Chip (MPSoC) and low-power embedded systems, including new thermal-aware design for 2D and 3D MPSoCs, design methods and architectures for Internet of Things (IoT) and wireless body sensor networks, dynamic memory management and interconnection hierarchy optimizations. In these fields, he is co-author of more than 250 publications in prestigious journals and international conferences, several book chapters and seven U.S. patents.
He received an ERC Consolidator Grant in 2016, the IEEE CEDA Early Career Award in 2013, the ACM SIGDA Outstanding New Faculty Award in 2012, and a Faculty Award from Sun Labs at Oracle in 2011. He has also earned two best paper awards at the VLSI-SoC 2009 and CST-HPCS 2012 conference, and five best paper award nominations at the DAC 2013, DATE 2013, WEHA-HPCS 2010, ICCAD 2006 and DAC 2004 conferences. He serves or has served as associate editor of IEEE Trans. on Computers (TC), IEEE Design & Test of Computers (D&T), IEEE Trans. on CAD (T-CAD), IEEE Transactions on Sustainable Computing (T-SUSC) and Elsevier Integration. He was the Technical Program Chair of DATE 2015 and General Chair of DATE 2017. He is also currently President of IEEE CEDA in the period 2017-2018, and was GOLD member of the Board of Governors of IEEE CASS from 2010 to 2012. He is a Distinguished Member of ACM, and an IEEE Fellow.