From the Digital-Twin to the Cyber Physical System Using Integrated Multidisciplinary Simulation: Virtualization of Complex Systems

Introduction

The pillars of multibody simulation, namely real behavior prediction, time-to-market and cost reduction, innovation, risk avoidance related to physical test have received an endorsement from recent digitalization developments and continuous increasing worldwide competitiveness. This has boosted the utilization of virtual simulation tools and increased collaboration between the virtual world and the physical world, with the goal to improve the quality and productivity, without neglecting efficiency and accuracy.

For achieving and maintaining competitive advantages, technology needs innovation: it implies the adoption of a digital process in the design cycle of product to investigate virtually its development, feasibility, performance, durability, cost, and manufacturing. This ambitious goal requires multidiscipline integration, co-simulation technology, optimization, data and process management, improvement of the correlation between physical and virtual tests. This means the adoption of the so-called Digital Twin, for integrating different simulation disciplines, tools and methodologies to better estimate system performances with respect single feature tools and including smart features like cyber-physical systems, data acquisition and management, human-machine interfaces. The Digital Twin, properly developed, adopted and installed, allows a full and direct control of the simulated process through its monitoring and, combined with simulation model estimation, improves and optimizes the full process, modifying its parameters, anticipating and predicting loss of efficiency and performances during production and operating life, optimizing maintenance actions. The accuracy of a virtual model is related to the collection, identification, and prediction of working conditions, but requires the right and acceptable compromise between accuracy and efficiency.

Other aspect to consider, to optimize, to simulate, to predict in a design, is the energy consumption and simulation tools can help to evaluate energy efficiency, through a right combination of empirical data/information and mathematical model, and – again - this is becoming relevant especially for real-time models for monitoring application.

Same importance has the estimation and prediction of end of life of components, generically evaluated using rough calculations, which could be too conservative or inaccurate.

A model-based approach has the ability to incorporate physical understanding of the monitored system, but it sometimes requires a very complex, difficult modelling activity, and it can be difficult, if not impossible, to catch the system’s behavior: many users could be skeptical about estimation using this method. Instead, nurturing the model using real data from sensors increases the capability of understanding and predict degradation’s behavior. Since they are based on real monitored data, current component degradation condition can be estimated. On the other side, some difficulties related to the lack of efficient procedures to collect and use data could require need to be managed efficiently.