DT consists of three elements: intelligent process control (IPC), AI_driven modelling and Decision Support Systems (DSS)

The combined use of simulation (finite element analysis, FEA) and data analysis reduces costs and increases reliability for quality predictive models.

Using simulation you can understand the production process, identify significant process parameters and the corresponding intervals. Thereafter, you can implement targeted experimental design of experiments (DOEs) and collect the data needed to build the models at a lower cost.

By applying multi-variable analysis techniques you can create accurate predictive models and validate them. The models can then be integrated onto the machine by creating virtual sensors or suggesting the correct process parameters to the operator.

Data-driven Digital Twins:

• support different tracking systems to identify each component and position, as well as communication protocols with process machines and quality control systems
• integrate multi-resolution and multi-variable process data monitored and gathered by a network of sensors across the distributed control system, and advanced models that link process variables to zero defect manufacturing(ZDM)
• activate real-time, knowledge-based reactions to any process variations and quality risks,
• implement automatic updating of process meta-models based on intelligent learning methods with virtual and real-time production information,
• calibrate direct measurements and their correlation with the process and target functions of the product, generally adapting the product requirements to control the level of uncertainty for significant parameters of the manufacturing process,
• display real-time data processing, safety messages and statistics production diagrams for multi-user interfaces such as machine operators, production managers, quality engineers and plant directors.

Designing automatic machines and robotic cells with Virtual Commissioning

Virtual commissioning is applied to the integrated design of mechanics, robotics, and automation of machines, automatic lines, and systems for factory logistics.

The 3D CAD model of the automated machine or production system is used to create a virtual commissioning model, integrating the programming logic of plc and robot controllers, to provide a realistic simulation.

These models allow you to design the system and validate the automation (interaction between sensors, actuators and control software) before its realization, with the aim of simulating its real behavior and preventing problems at the time of commissioning (a "marriage" between the physical machine and the control system). In addition, augmented reality features allow the use of virtual commissioning models to support the sale and maintenance of automatic machines, and for simulation-based training.

Design and optimize systems and production lines with discrete event and agent-based simulation

SIMUL8 is a simulation software for modeling, analyzing and optimizing the system-level performance of production systems during their design, reconfiguration, and production planning phases.
The simulation building blocks allow you to create accurate system models of complex system architectures, such as production lines, job shops, robotics cells, assembly systems and complex product flows. Creating a preliminary model requires limited information (for example starting from a process list and the 2D layout of the plant), while it is possible to progressively add detail to the simulation.
SIMUL8 models provide decision makers with critical information, such as:

• resource utilization levels (machines, operators, etc.)
• buffer saturation
• the flow of materials and semi-finished products
• the efficiency of the production system

The models can be easily connected to optimization algorithms by considering the output objectives (throughput, quality yield), and constraints (machine capacities, failure rates, shift patterns, and other factors affecting the total performance and efficiency of production systems).

Therefore, they allow you to evaluate the effects of different scheduling policies, batch and production flow management and operators, maintenance policies, the impact of system modification interventions (modification of the layout, replacement of machinery, interventions aimed at reducing the average return time of certain events, etc.), the effect of adding new products online, etc. and to identify the best compromise solutions in the face of conflicting objectives (e.g. costs vs. productivity).

Develop asset-level condition monitoring and predictive maintenance models

The combination of technologies for numerical and symbolic calculation (Maple) and a multi-domain simulation systems based on Modelica (MapleSim), allows you to create efficient simulation models that integrate mechanical, electrical, hydraulic, etc. components and their controls. Based on the ability to import 3D CAD models it is possible to accurately model articulated systems (kinematic chains) taking into account masses and inertia. The combination of these features enables accurate modeling that emulates the behavior of real systems.

The efficient generation of C code according to the functional mockup interface (FMI) standard allows you to export the models in the form of a functional mockup unit (FMU) and to support both model exchange and co-simulation by integrating other technologies. The resulting models can be integrated into the machine in the logic using software in the loop (SIL) by creating a digital twin for asset condition monitoring and predictive maintenance.

As the product moves from design and manufacturing into operations, simulation can continue to play a pivotal role in delivering the best possible results in the field.

By using remote sensors to gather data on a product’s working conditions, analysts create a virtual replica — a digital twin — of that product and then apply the same physical forces and other environmental conditions to the digital model. Applying simulation as part of a digital twin can provide vital insights in the form of virtual sensors, in situations where no physical sensor exists or would even be possible. Simulation can also run what-if studies for optimal performance and can predict critical failure or maintenance requirements.
Simulation is also increasingly applied to the manufacturing phase, where it significantly improves the efficiency, cost-effectiveness and flexibility of production. With the rise of mass customization of products — made possible by additive manufacturing, or 3D printing — simulation helps ensure that the finished product has the optimal shape and is made accurately, cost-effectively and with a high degree of consistency over time.