In this plenary session, the importance of utilizing big data analytics to improve process safety, reliability, and productivity will be highlighted. Presentations will cover a wide spectrum of topics from approaches that improve decision making on process safety to how process industries (such as chemical, oil & gas, semiconductor, and food) find actionable insights from Big Data to optimize process operation, and therefore, increase plant reliability and integrity. Big data can be identified by the four Vs (Volume, Velocity, Variety, and Veracity). With the continuing increase of plant data (Volume), there is a need to identify approaches that efficiently bring relevant information from data to improve process safety (for example, accelerating investigations into the root causes of safety incidents utilizing big data analytics). Speed in data preprocessing and analysis (Velocity) contributes to actions in real-time that avoid disruptions in production; novel approaches to identify process safety indicators that proactively anticipate safety abnormalities are an area of interest for this session. Contributions that benefit from utilizing data from various sources (Variety) and create an impact on decision making, development of leading indicators and Hazard Risk Assessment are also welcome in this session. Examples of data sources include raw materials, process operation, product quality, customer feedback, and other unstructured sources (i.e., text data, images). Finally, big data can have biases and inaccuracies (Veracity) that need to be identified and accounted for in order to reduce inference errors and improve the accuracy of generated insights.
Data and Analytics for Human Health at NASA
Robert J. Reynolds, Mortality Research & Consulting, Inc., National Aeronautics and Space Administration
Human spaceflight presents a unique set of hazards to human health, necessitating ongoing studies and surveillance of astronaut health. However, in spite of almost 60 years of spaceflight experience, from an epidemiological perspective there has been relatively little human exposure to space – particularly over long durations. This sparse data environment makes learning from observational spaceflight data challenging, and requires the use of data from various spaceflight analogs as well as non-traditional methods of analysis. By methods such as analyzing intermediate outcomes along a causal pathway, formulating alternatives to null-hypothesis statistical tests, and utilizing machine learning algorithms, researchers continue to grow our knowledge about the long and short-term effects of spaceflight.
In this talk we will review the amounts and types of human health data available at NASA, and discuss several examples of analytic techniques employed in its data-sparse environment. We will examine the strategy behind health surveillance, decision making tools, and the general epistemological approach behind contemporary health research for spaceflight.
Transforming the Process Industries Through AI and Digitalization
Antonio Pietri, CEO, AspenTech
The process industries are faced with an increasingly volatile, uncertain and competitive business environment, with new and changing requirements, regulations and opportunities coming at breakneck speed. At the same time developments and trends in AI and digitalization (e.g. connected devices, IIoT, the data explosion, Cloud and Edge computing, artificial intelligence and machine learning) are providing new opportunities to react effectively. In this presentation, Antonio Pietri will discuss examples of how AI and digitalization are being combined with domain knowledge to transform process engineering, production optimization, and plant reliability. He will also present his vision of the “Smart Enterprise” where plants will become increasingly self-learning, self-adapting, and self–sustaining.
Are you looking for more in this area? Check out these sessions:
Industry 4.0 & Fuels and Petrochemicals Division Joint Plenary, featuring from Lloyd Colegrove, Director of Data Services & Fundamental Problem Solving at Dow and Max Eklund, Business Incubation, Technology Liaison, Fluor, speaking on the current transformation of the chemical industry.
Panel Discussion: How Can Industries Engage in Data Science Education in Chemical Engineering? including perspectives from Georgia Tech, University of British Columbia, University of Washington, and Auburn University.