GET THE APP

Predictive analytics for the process industry goes far beyond ass | 17551
Journal of Chemical Engineering & Process Technology

Journal of Chemical Engineering & Process Technology
Open Access

ISSN: 2157-7048

+44-20-4587-4809

Predictive analytics for the process industry goes far beyond asset monitoring. Is your process historian still up for the task?


International Conference on Chemical Engineering

September 12-14, 2016 Phoenix, USA

Hans De Leenheer

TrendMiner, Belgium

Posters & Accepted Abstracts: J Chem Eng Process Technol

Abstract :

For years now we have been searching for the holy grail of analytics in the process industry. Every big data solution has tried to crack the key to unlock predictive and prescriptive analytics, having the system to tell you what you should be doing. Very rarely we have seen success here. The only place where we have some success is where a very specific question with a specific resolution can be repeated indefinitely. This could work when you have thousands of identical assets, like pumps or windmills. But this does not work for the chemical process. In the chemical industry, engineers deal with dozens of different processes at a time, each reacting differently to internal and external variables. This makes repeating an exact question to an exact solution impossible. For this we will need a completely different approach. This is also the very reason why this cannot be done by a data scientist but by the engineer himself. In this presentation, I will lay some foundations on why existing big data solutions will not suffice for the process industry, why the data historian is not the place to be and why the data scientist is not your next new hire. At the end of this presentation, I will show a short live demo of TrendMiner, one of the new solutions that will help the process engineer get the information out of his data.

Biography :

Email: hans@trendminer.com

Top