AI-Based Simulated Learning For Staff Training And Plant Optimisation
Training specialised personnel in the handling of complex plants is not a trivial task. With “immersive training”, this can be realised much better in a kind of game-like approach. It is a process of learning using a simulated or artificial environment. The environment allows learners to be fully immersed in learning in a way that makes them feel like they are experiencing a real learning environment.
Aveva, a company based in Sulzbach, Germany, has developed ‘ITS’, a process that combines a true-to-life process simulator with a virtual walk-through environment of a plant (a so-called digital twin). Via this, every action in a virtual environment can be set to trigger the dynamically correct reaction in the plant in real-time.
Immersion in applications is created with the help of xR technologies – augmented reality, mixed reality and virtual reality. In this context, this is often referred to as gamification, as the learning process actually contains something playful – you can ‘play around’ with the digital twin until you go to the real system.
It is also important to distinguish between interactive and immersive, which is often confused. The difference between immersive and interactive is that immersive means to be truly immersed in something – while interactive expresses to act with each other (persons and objects).
Advantages Of Simulated Learning
AI-driven simulators reduce training time for new equipment, improve cost efficiency and thus optimise the return on investment (ROI). Data collected by Aveva shows that, depending on the company, such a system can cut costs by 30 to 40 per cent, reduce recovery times from shutdowns by 15 to 20 per cent and reduce maintenance budgets by one to three per cent.
‘In certain industries, the benefits simply can’t really be put into numbers anyway – for example, the avoidance of human error in the operation of nuclear power plants,’ emphasises Maurizio Gatardo, CTO of the Hessian branch of the software company with its headquarters in Cambridge (UK).
The real challenge in installation is usually that there is a lack of data. If you want to create a digital twin, you have to have 100 per cent of the construction data of the buildings and installations for the feed. However, not all companies have this information, as they have grown organically over a long period of time and corresponding documentation is sometimes patchy.
Once this hurdle is overcome, the benefits of simulated learning are obvious – humans and machines can interact recognisably better in future.
Photo: Monopoly919