Simulation software such as industrialPhysics acts in the engineering process as a test bed for the design of the artificial intelligence (AI) that is to be developed, because with the aid of the digital twin a virtual training centre is created for the purpose of preparing the algorithms for your task: namely monitoring the productivity of machines and plants. In principle the simulation is currently used by engineers as a tool with which to obtain findings that could be transferred to reality in accordance with the directive VDI3633-1. In the continuation of this approach, the engineer exposes the AI to its learning task in the simulated environment prior to the operation of the real machine.

Hardware-in-the-Loop simulation and AI

The training environment, based on a Hardware-in-the-Loop (HIL) simulation, consists of the machine model, a real or virtual controller and the AI system. The interaction of all three components can thus be tested. This is important in order to enable the artificial intelligence to test various algorithms and experience possible errors itself, with the aim of avoiding them in real operation. The learning goals can thus also be defined and checked in this way by the engineer without risk.

Ghost robots

In many branches of industry that work with robot-controlled processes, various robots stand together in a confined space and must function safely next to and with one another. The more the self-programming machine becomes reality, the more thoroughly one has to check for collisions or poor reachability. A real-time 3D model with collision detection can be used for this.

A ghost robot runs in parallel with the virtual machine and ahead of the real machine by a predefined time sequence. With the help of this additional tool it is possible even during the planning and development to run virtually through various scenarios of how the robots concerned can best operate with and next to one another. In addition, the simulation of the robot behaviour helps companies to comply with the Machinery Directive. Hence, the speed of the movements can be adapted to the environment, e.g. near to scaffolding or protective fences, at an early stage of the development process.

Machine Learning

The use of the 3D-simulation as a training environment for machine learning leads the engineer directly to the next evolutionary step: the fusion of both technologies in real operation, because the degrees of freedom of machine learning – and thus the bandwidth and computing power required later on in operation – can be reduced considerably through the use of a 3D model of the machines and robots. That will be of key importance when all machines in the production plants are enhanced through AI. The cost-effectiveness of the computing power provided rises and falls with the number and complexity of the degrees of freedom to be learned through algorithms. The 3D real-time simulation with collision calculation on the machine can help to considerably reduce the degrees of freedom to be recorded and the AI can thus focus on the essential things.

Would you like to learn more? Just contact us, we'll be glad to advise you!

More information

Related topics and services

Ready to start smart engineering?

Request more information about our simulation software industrialPhysics.

Request information

User login

Enter your username and password here in order to log in on the website:

Forgot password?

Not yet registered?

Request Information

To provide you with targeted information, such as industry examples, it would be helpful for machineering to answer the following questions.

* Information required to process your request. More information about our privacy policy can be found in our imprint.