Machine Learning in the NAPCON Advisor
News / 12 Nov 2019
The newest member of NAPCON Understand is our top-notch digital operator assistant, NAPCON Advisor. It is currently in pilot use with our partners and will be released during 2020.
As a digital operator assistant, NAPCON Advisor is an intelligent software solution that supports panel operators in their daily work.
- Its goal is to help run the entire processes optimally.
- It enhances proactive operation, instead of reactive and alarm driven.
- It supports learning by doing.
- It is based on machine learning and optimization.
The starting point for a digital assistant is the definition of optimal process operation. In a steady operation, a digital operator assistant takes into account aspects of safety, economics, stability, and operability when advising the panel operator. During transients, the objective is to return to a steady operating mode as soon as possible, simultaneously minimizing the amount of off-spec production and harms for other units.
With the dynamic machine learning process model of NAPCON Advisor, you can either predict the process behavior, run “what if”-scenarios to support the decision making or let the advisor suggest operator actions based on optimization.
Take the profitability of the refinery to the next level
It is possible to take the profitability of a refinery to the next level by utilizing artificial intelligence (AI) and machine learning (ML). The objective of the usage of ML and AI is to refine data and turn it into valuable information, knowledge, and insight that helps either to run the plant in a more optimal way or to detect and identify abnormal events.
Today, it is imperative to operate refinery units in an agile way. The refining industry is experiencing two trends. Firstly, heavier crude oils need to be processed, and, secondly, different bio-based feedstocks are used more and more. As a result, processes are more often in a transient state, and optimal process conditions may be significantly different from the ones that the industry has been accustomed to. At the same time, a new generation of process operators is entering the labor market. All this calls for redefining the way processes should be run and how operators could be assisted in performing their tasks optimally.
AI can be a beneficial tool in this
AI can learn the best practices to operate processes in various situations and can help operators achieve their targets. This type of digital application builds a bridge between traditional paradigms of process modeling and competence development. Our first experiences on ML-based digital process operator assistants are very promising, and we believe that they are going to play a significant role in panel operator toolsets.
Learn more in the white paper: Machine Learning in the NAPCON Advisor