Plenaries

As technologists, we often focus on developing new technical capabilities without thinking about the potential consequences or societal implications of these technologies. Instead of solely asking ourselves how to create a solution, we should also be asking ourselves about the potential impact of our developments. This talk will explore this notion by examining valuable lessons learned from the rich, yet complex, history and evolution of artificial intelligence.

Abstract: The deployment of critical autonomous systems has changed the world and improved the lives of many, from satellite systems such as GPS to Amazon Robotics deploying over 500,000 robots into production, in many ways laying the foundation for how mobile robots will evolve and be safely inserted into our society.  

In this talk I will present a glimpse of the development efforts in 1) model predictive control technology and 2) machine learning and statistical process analytics that are related to my personal academic experience. Then a new framework of dynamic latent variable (DLV) analytics is presented in detail with an application to process troubleshooting. I will share research and collaboration experience with industry and via the academia network to highlight the following aspects: i) there have been vigorous interactions between industry and academia; ii) human acceptability and interpretability are a must for new technology adoption, including machine learning; iii) innovative theory and solutions should have real-world needs; and iv) domain knowledge is necessary for machine learning and data analytics.