ISA Interchange

Welcome to the official blog of the International Society of Automation (ISA).

This blog covers numerous topics on industrial automation such as operations & management, continuous & batch processing, connectivity, manufacturing & machine control, and Industry 4.0.

All Posts

Why and When to Use Machine Learning for Automation Applications

Introduction

I see machine learning (ML) as the last and most complex resource when developing a process model used on a specific application (for instance, process control) to estimate a process variable value (also known as soft sensors) or to predict failures. The key point is that there should be a proportionality criterion between the engineering solution and the problem to resolve. While overdesigning takes more effort, time, and money, nonetheless the achieved benefits are not in the same order of magnitude as the engineering labor. Depending on process complexity and model performance requirements, modeling techniques vary from simple to elaborate.

Modeling Techniques

First, let's define a complex process/phenomenon: A system that heavily depends on initial conditions, is non-linear, has some behaviors that are stochastic, and has the sum of its parts bigger than the total.

For well-known deterministic industry processes, models can be obtained by applying first principles. However, due to model uncertainty, or modeling simplification (e.g., linearization, approximations) the capability to emulate the process, as per required performance, is negatively affected. To compensate this gap, the model can be completed by adding a function attained from data regression and/or a probability distribution function.

When the aforementioned attempts are not capable to produce a suitable model per application requirements, a more sophisticated method is necessary. Here is where ML is an option to consider. Of course, it has to be justified. In other words, it has to be feasible to develop a model and the potential benefits need to be reasonably achievable.

Machine Learning (ML)

ML is a useful model built from data. There is a dataset to train the model and a separate dataset to test the model. When the model is deployed, it is intended to replicate the process, even if the inputs values are different from those used to train the model. In other words, it is expected to generalize with a pre-established performance accuracy. As ML models can be affected by prediction errors or overfitting, model parameters can be adjusted to improve its performance.

Why Should ML Be Implemented?

As mentioned, if with conventional techniques (i.e., first principles, linear regression, or statistical models) a suitable model cannot be obtained to fulfill application requirements, then ML is an option that can be assessed.

ML could be seen as a black box, as its implementation is in the form of programing code (e.g., Jupyter Notebook) or connecting diagrams (e.g., Microsoft Azure). If one reads the coding, they will not be able to directly read the model structure (input and output variables, how they are correlated, the constraints, etc.), whereas in conventional modeling and programming, model setting can be understood with minor effort.

How Can ML Be implemented?

Here are some examples of potential industrial applications for ML:

  1. Virtual Analyzers (Soft Sensors). These are intended to determine a physical property of the process fluid based on operational data (process variables, operational modes, etc.). This approach is a suitable option whenever there is no commercial analyzer for a specific physical property, or those available in the market are not able to measure the property as the measurement principle is negatively affected by operational conditions. On the other hand, it is desired to avoid all the cost (installation and maintenance) associated to an analyzer and have a virtual instrument that could achieve this with reasonably accurate measurements.*
  2. Predictive Maintenance. This model can alert when an equipment will require a special action, besides regular preventive maintenance tasks.
  3. Predictive Models. For process control, optimization, demand forecast, etc.*

Conclusion

Engineered solution complexity increases with problem complexity, ideally in a linear proportion. Albeit it might be exponential for some particular cases. ML could be overkill for applications where conventional techniques are suitable, as ML additional effort can only deliver the same benefits as conventional techniques. Any applications has to be assessed on a case-by-case basis.

 

*Note: These instruments can also be developed based on first principles, therefore achievability and benefits will dictate where to go.

Angel Prieto
Angel Prieto
Angel Prieto is an Electrical Engineer by training with a MSc Engineering degree in Electronic Engineering, who developed a career as an Instrumentation, Process Controls and Automation Engineer, with more than 20 years of experience in Multidisciplinary and Automation Project, Technical Support and Operations Engineering in the Oil and Gas Industry. Mr. Prieto is also a Professional Engineer (PEng.) for the Provinces of Alberta and Saskatchewan in Canada, as well as an ISA Certified Automation Professional (CAP).

Related Posts

What are the Biggest Mistakes You Have Seen? Part 3

The following discussion is part of an occasional series showcasing the ISA Mentor Program, authored by G...
Greg McMillan Jul 1, 2022 5:30:00 AM

How is Automation Changing Chemical Plants?

Automation is becoming a gamechanger in various industries. Decision makers need to figure out how best t...
Emily Newton Jun 28, 2022 5:30:00 AM

ISA Commemorates International Women in Engineering Day

Since 1982, women have earned almost 10 million more college degrees than men. However, despite this stat...
Steven Aliano Jun 23, 2022 5:30:00 AM