The availability of extensive sensing capability and ubiquitous computing has led to numerous technological breakthroughs, from smart gadgets and personal assistants to Internet-of-Things and personalized medicine. In addition to their traditional roles in industry and consumer applications, sensors are nowadays embedded in numerous systems to provide a better understanding of the ambient conditions. In many cases, these sensors are nearing their limits of theoretical performance. Nonetheless, there are many existing and upcoming applications that require more accuracy, precision, or information. The current trend is to take advantage of the available computing power at different levels (Edge, Fog, and Cloud) to improve some quality metric(s) of the measurements on parameters of interest. The available signal processing algorithms include the traditional techniques such as filtering but go well beyond them to not only produce numbers but also understandings of the environment.
The aim of this session is to provide you with the basic understanding of techniques that can be used to improve the quality of data collected by sensors or to gain additional knowledge from sensors. The discussion will have the following components:
- Measurement of physical and chemical quantities.
- Basic transduction principles and common interfaces.
- Noise and non idealities in measurements.
- Optimal method to combine signals from similar sensors.
- Kalman filter for sensor fusion and noise reduction.
- Continuous models (i.e., regression) for linear and nonlinear sensor outputs.
- Algorithms for discrete models for sensor output (i.e., classification), including k nearest neighbors, support vector machines, logistic regression, and neural networks.
Basic knowledge of electrical and mechanical measurements; elementary statistics; basic system theory; introductory linear algebra.
Lecture slides will be printed and provided to the attendees. Sample datasets and codes for the material covered in the course will be provided to the attendees as download links (will require Matlab or Octave).
Engineers and engineering students that use sensors or sensor data in their profession. The topics are applicable to different fields of engineering from electrical and mechanical to chemical, structural, and computer engineers.
Associate Professor, School of Mechatronic Systems Engineering, Simon Fraser University
Associate Member, School of Engineering Science, Simon Fraser University
Behraad Bahreyni is an Associate Professor and the founding Director of the Intelligent Sensing Laboratory (ISL) at the School of Mechatronic Systems Engineering at Simon Fraser University, BC, Canada.
He received his B.Sc. in electronics engineering from Sharif University of Technology, Iran, and M.Sc. 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 NanoScience Centre at Cambridge University, UK, where he conducted research on interface circuit design for microresonators. He joined Simon Fraser University in 2008 after a 1-year tenure in the industry as a MEMS design engineer.
Over the past decade, his research activities have focused on the design and fabrication sensing systems comprising micro/nano sensors from silicon, polymers, or nanocomposites, their interface electronics, and the required signal processing algorithms. Dr. Bahreyni is the author of more than 100 technical publications including a book on the fabrication and design of resonant microdevices.