Dear Colleagues
In recent years, Artificial Intelligence (and Machine Learning) has been touted as the panacea for process plants: using instrumentation to gather the data and automated systems to effect improvements to the operations. There have been huge advances and some great successes, but maintaining a healthy cynicism is essential.
Machine learning is certainly one of the most exciting technologies in recent times. Both large and small companies have embraced it with tremendous results.
There are three major types of machine learning:
Supervised learning, or the relationship of given inputs to a set of outputs. For example, the sensor inputs in a process plant (using previous breakdown knowledge) are ‘wired’ to predict the likelihood of the failure of a pump, or a maintenance prompt for equipment.
Unsupervised learning is where no pre-defined relationship exists between input data and an output variable. For example, a sensor is programmed to log data from hundreds of similar industrial plants, searching for specific patterns for optimal operations. The algorithm can trawl large data sets and then classify them.
Reinforcement learning uses algorithms to perform tasks, but they are designed to become ever more astute as feedback is received. They strive for actions that are rewarded and then optimize themselves in response. They are effective in a scenario where little training data exists and there is no clearly defined end state. For example, balancing the load on electricity grids in varying demand and supply cycles. (This is especially relevant these days with the growth of smart grids.) Or the optimization of self-driving cars.
A fourth machine learning technique is deep learning (using ‘neural networks’). Impressive advances have been made in image processing and recognition, robotics, and natural language processing.
Are these advances achievable for industrial plants and industrial automation? The picture is more nuanced than the AI consultants would have us believe. There are examples of singular successes: with chess games, machine vision, voice commands, and analysis of text data. These commercially useful applications have lured us into the belief that AI can be ably applied to industrial plants. Almost superstitious awe exists around the capacity of AI.
The widely quoted industrial IoT guru, Jonas Berge, suggests caution. For instance, a dubious claim is that an analysis of existing plant data (such as the correlations between flow data and pump failures) will provide new insights and ultimately optimize plant processes. Often this costly exercise merely reveals what you already knew or suspected.
There are some practical approaches to implementing these technologies. When data is collected from instrumentation, filter out and separate the ‘noise’ from the real data. Spurious, but seemingly important signals can be picked up. Instrumentation specialists are critical here; the data needs to be assessed by them before further analysis is completed by those with broader process knowledge.
It is also important to bear in mind that events, such as the failure of a conveyor or pump, may occur after a number of years rather than months. Hurried data collection can prove inadequate and skew results. And ensure that an appropriate number of sensors are in place.
As opposed to complex machine learning, a simple rules-based system is best when analyzing plant data. A flow sensor, for instance, is level-sensitive; it acts by setting off an alarm or shutting down a pump, or closing a valve.
It has been suggested that AI is best-placed to analyze humans; to find useful patterns from a morass of complexity. Machines, on the other hand, are predictable; it is harder to gain value from applying AI and machine learning to them.
In conclusion, instrument specialists remain critical to industrial plants.; their deep knowledge of the processes and operations ensures they run smoothly. Yes, AI is producing innovations, but not in terms of having a genuine and useful impact on plants ……just yet.
Yours in engineering learning,
Steve Mackay – PhD
Dean of Engineering
