Abstract: Technological advances have provided abundant computing power, widespread connectivity, and various microsensors to quantify real-world parameters, leading to revolutions such as the Internet of Things (IoT) and wearable technologies. Increasingly, recognizing the context in which data is generated is more relevant than the raw sensor data itself. While sophisticated machine learning algorithms can process raw data to generate context, the ability to generate and record data has surpassed the capacity to make sense of it efficiently.
The next paradigm in sensing systems is to integrate sensing and cognition at the sensor level, enabling pattern detection in data as it arrives without the need for extensive data transmission. This talk covers three such physical computers based on physical Reservoir Computing (RC) to detect patterns in data. First, an electrothermal computer is built using off-the-shelf components to tackle sophisticated contextual computations, including standard benchmarks and real-time event detection. We then demonstrated similar capabilities based on 3D-printed computers for near-sensor signal processing. Finally, we will discuss the evolution of the work into building a 3D-printed computer with concurrent sensing and computing capabilities as a demonstration of in-sensor computing. We conclude with potential roadmaps for future research and a diversity of applications that can benefit from its results.
Biography: Dr Behraad Bahreyni, SMIEEE, PEng, is a professor and the Director of the Intelligent Sensing Laboratory at Simon Fraser University, BC, Canada. He received his BSc from Sharif University of Technology, Iran, and MSc and PhD degrees in electrical engineering from the University of Manitoba, Canada, in 1999, 2001, and 2006, respectively. He was a post-doctoral researcher with the NanoSicence Centre at Cambridge University, UK, where he researched interface circuit design for microresonators. He joined SFU in 2008 after a one-year tenure in the industry as a senior MEMS design engineer. In 2016, he was with NXP Semiconductors, the Netherlands, developing advanced signal processing methods for sensing. His research activities are focused on the design and fabrication of micro/nano-sensors for applications that include consumer electronics, mining, automotive, and space exploration. Dr Bahreyni has contributed to more than 150 technical publications.