Most people know about the first industrial revolution. That's the era when there weren't enough strong humans to make the products and infrastructure society was asking for. The invention of the steam engine came at the right moment: muscles were replaced by steam power. Today, we face another shortage: a growing number of open "engineering" positions can be seen, and they stay open for longer as well. As processes and installations become increasingly advanced, it's increasingly challenging to design them and keep them in top shape.
The solution for this? We'll again see a sort of "replacement". At this time, technically skilled minds with a strong touch of creativity will be replaced: engineers. We're not referring to Chat GTP or Dall-E (this actually is our first reference to these products, so we're obviously strongly lacking some skills in following LinkedIn posting trends). A virtual engineer will not be too different from a "human" one, as it will be based on physics, but it will also include AI-based functions and capabilities. In the current piece, we focus on those engineers taking care of machines and other critical installations in manufacturing, not those designing and building bridges.
The change upon us?
What is new is that this approach will lead to new ways of structuring even a company. Before, different departments took care of individual tasks: energy efficiency, product quality, maintenance... Using a virtual engineer, these aspects will be combined, as they are closely linked. For example, decisions focusing on product quality, like changing process temperature, often have an impact on maintenance needs and energy usage, or remaining lifetime. Currently sometimes the data sets required to achieve this are too large, or the situations too complex for the human mind to handle, so a helping hand could really be beneficial here.
Virtual Engineers will give birth to new roles, making operations more efficient and sustainable by combining multiple activities that are today still located in different departments
Not what you would think...
A "virtual engineer" might not be what you would expect. It's not going to be a humanoid robot walking around the mill wearing an overall and hard hat. The first ones will be purely software-based. Where the human one relies on eyes and ears, the "virtual" one will rely on input from different kinds of sensors. Sometimes video, but most often raw data streams of pressures, vibration signals, temperatures, deformation signals, positions, noise... This is complemented with a set of working principles and metadata, such as what the machine looks like, the components it consists of, how it is constructed, what it is making, and how...
It's all about combinations
How can a virtual engineer do all of this? By combining physics and advanced ways of treating large data sets, let's say: AI... Physics is used as a basis for the "limits" of operations. The AI aspect, on the other hand, is used, among other things, to learn (and use) what is normal behavior for a type of machine, what is usually happening after a type of event or operation. From the moment the "engineer" is applied to an individual machine, it will also learn about the specificities of that machine, making it even more specific and sensitive.
... and limitations
There are, of course, limitations to all of this. Today, we cannot yet expect that such a virtual engineer is able to do what a human one is doing: it will not be able to solve somewhat all problems on multiple machine types. What is possible is to have a specific virtual engineer that can take up this role for a specific type of machine or installation. Each will get better in doing their job related to that specific type of machine. A human engineer, at university or engineering school, learns a lot of physical principles as well as methodologies to approach a problem or situation. These methodologies can be programmed in as well. It will involve looking at all the relevant data, including the context of that data and its origin (where in the installation, what mode of operation does it relate to...) and, after proper analysis, include more data or draw conclusions from it.
So... today, we witness the birth of the concept of virtual engineers. They are by far not (yet) the full equivalent of a human engineer, but will be able to take up this role for specific applications and installation types. Moreover, considering the breadth in data streams crunched by these virtual entities, they will give birth to new roles, making operations more efficient and sustainable by combining multiple activities that are today still located in different departments.Vir