ALT + + Change font size
ALT + / Change contrast
ALT + Q Quicklinks
ALT + P Header menu
ALT + M Main menu
ALT + U Footer
ALT + G Sitemap
ALT + O Search
ALT + I Content
ALT + K Contact
ESC Reset All
X

Research

The VAL undertakes application-oriented research in the following fields:

 

Further information on the individual research fields and our current research projects is available here:

 

Digital Twin

Digital Twin


The core competence of the VAL is the investigation and utilisation of Mixed Reality (MR) methods in mechanical engineering. Virtual commissioning, training and monitoring are just some applications, by way of example.

The VAL is successfully developing a novel framework to facilitate the engineering of a 3D scene in the standard Web browser and the automatic generation of 3D holograms on various terminal devices, such as Augmented Reality (AR) glasses (via Optical-See-Through) or tablets (via Video-See-Through) and Virtual Reality (VR) glasses. The hologram on the terminal device can be linked to real sensor data via the edge cloud to move the hologram of the installation in real time, for example, or to superimpose status data at relevant locations.

Human-Machine Interaction

HUMAN-MACHINE INTERACTION


Mixed Reality (MR) is a method which is becoming more and more important in the context of digitisation and the Smart Factory. These modern methods of visualisation not only facilitate the immersive and three-dimensional visualisation of virtual content, they also make new forms of Human-Machine Interaction possible.

The VAL focuses its attention on the investigation and integration of the device-independent (Augmented Reality, Virtual Reality, tablet, smartphone etc.) and platform-independent interaction/visualisation of the Mixed Reality in the Loop simulation model. The visual interaction with a human being involves studying and observing both ergonomic as well as technical criteria, and also giving due consideration to general concepts such as design laws and perception. The VAL also investigates and implements new possibilities for Human-Machine Interaction such as X-ray vision into the machine and production in slow motion. A further focus of the research work is the direct, intuitive and natural real-time manipulation of the interactive simulation model, via gesture control for example. Another topic of interest here is the sharing of a MR scene with several observers/actors (Shared Experience), so that several users can interact in a shared virtual world.

Artificial Intelligence

Artificial Intelligence


Artificial Intelligence (AI) methods offer great potential for small and medium-sized enterprises (SME), and in the future they will play their part in sustaining Germany as a centre of commerce and industry. The challenges facing SMEs when using AI can currently be identified as the configuration of existing AI algorithms and also the lack of reusability.

The advances being made in computer infrastructure, availability and range, as well as the development of dedicated hardware, mean that bigger and more complex AI algorithms can be trained more quickly. The VAL is searching for solutions which can provide access to this next generation technology. The focus is on their application in the context of the Digital Twin, as this will allow live and simulation data to be used to develop and optimise AI algorithms. For low-threshold entry to AI, a service-based configuration system is being developed to help the user to create AI systems.

The research at the VAL concentrates on the following topics:

  • Speech input and output
  • Model-based object recognition of 3D objects and images
  • Configuration/author system for AI systems
  • Value-added services as AI enablers (e.g. data recording, cloud access,...)

Autonomous indoor aerial robotic systems

AUTONOMOUS INDOOR AERIAL ROBOTIC SYSTEMS


The consumer market is experiencing a growth in the demand for individualised products. This is accompanied by a demand for ever shorter product life cycles. Numerous developments and research results relating to flexible and ultra-flexible production in the context of a flexible factory show that the goal is for production plants to have a high degree of adaptability, dynamisation and autonomy. To this end, the process of adapting individual processing stations to a broad spectrum of production steps has to be flexible and automated, or individual processing stations have to be combined with each other to achieve a production step. A further requirement is that combinations and sequences of spatially separate process steps have to be developed further by means of flexible intralogistics and handling in order to meet the increasing demands.

The Virtual Automation Lab at Esslingen University of Applied Sciences is exploring the use of aerial robots (e.g. multicopters) to transport small parts between processing stations to increase the flexibility of production processes. The aerial robots allow the mostly unused air space in the production halls to be utilised and thus provide a high level of dynamism and flexibility. Using aerial robots in production halls involves numerous challenges:

  • Simultaneous indoor position detection of several aerial robots
  • Control and feedback control for autonomous take-off, flight and landing
  • Collision-free online trajectory planning
  • Multi-UAV control: control of several aerial robots in restricted spaces
  • Commissioning in changing environments

The Digital Twin of a production facility provides the spatial information including online updates of mobile and dynamic components and thus has a high level of informational content. The Virtual Automation Lab is investigating how this high level of informational content can be utilised in relation to control, online trajectory planning and indoor localisation. 

Indoor localisation

The increasing availability of machine data in almost real time which has come about through the fourth industrial revolution facilitates the continuous monitoring and determination of the machine status. By using the geometric representations of the machines, combining them with real-time machine data, and supplementing them with machine models, it is possible to create the Digital Twin of the production facility and use it to derive 3D maps of the production environment in real time. These are now available for the indoor localisation of the aerial robot.

The availability of the real-time model of the environment affords the researchers the opportunity to investigate new methods of indoor localisation which use computationally efficient and easy to install systems and also employ a minimum number of sensors. The indoor localisation has to master the following demands and challenges.

  • Use the Digital Twin with real-time data to generate 3D models of the environment and for the spatial orientation therein, paying due consideration to minimising the volume of data so as to reduce the data load on the frequency bandwidths of the wireless communication technologies which are being used.
  • Take the existing infrastructure into account: exploit the system-specific properties of wireless communication technologies (e.g. transmission power of WLAN Access Points) as input values of the localisation method.
  • Keep the number of sensors to a minimum, thereby producing a system which is low weight, low cost, energy efficient, and generates a low computational load.

Online trajectory planning

Using UAVs to transport parts on production lines harbours a high dynamic potential for collision when several UAVs are in this space and execute independent flight manoeuvres in parallel.

Localised curve flows, and a model of the environment from the Digital Twin which is updated in real time, are taken as the basis for fitting the flight trajectories to the current state of the environment at a given runtime by superimposing virtual forces such that they are always guided around static (e.g. building structures or utility supply pipes) and dynamic (e.g. other UAVs) obstacles (see Fig. 3).

A systems theory analysis of the trajectory behaviour (trajectory dynamics and stationary final position) facilitates the parameterisation of the so-called curve flow method (CFM) with the aid of physical quantities. A comparison with the Elastic Bands Method, which serves a similar function and is frequently used in mobile robotics, shows that the CFM has the following significant advantages:

  1. Numerically stable solvability
  2. Analytical parameterability
  3. Scale independence and
  4. A slightly better computational efficiency.

To validate the method, several instances of the CFM executed in parallel were put into operation each on its own industrial control systems, which are networked via a local cloud. The CFM was implemented as a separate (collision-free) point-to-point type of interpolation here so that sequential programs could be used to help it fly to various points in succession without having to explicitly take the collision potential into consideration in the programming.

The collision-free executability of the target trajectories generated by the CFM was verified by means of several test scenarios in the form of HiL simulations with dynamic models of the UAVs (see Fig. 4) and also with real UAVs.

Research projects with public funding

RESEARCH PROJECTS WITH PUBLIC FUNDING 


Service-based AI configuration support as an accelerator for AI applications in SMEs - accelerateKI

Artificial Intelligence (AI) methods offer great potential for small and medium-sized enterprises (SME), and in the future they will play their part in sustaining Germany as a centre of commerce and industry. The challenges facing SMEs when using AI can currently be identified as the configuration of existing AI algorithms and also the lack of reusability. This research project aims to increase the usability of AI algorithms by means of service-based AI configuration support and to thus systematically decrease the reluctance of SMEs to use them. 

The project is funded by Baden-Württemberg Ministry for Science, Research and the Arts.


Hybrid interaction concept to provide training by means of Mixed Reality in the Loop Simulation - MRiLS

The Virtual Automation Lab (VAL) at Esslingen University of Applied Sciences is part of a collaborative project which is aiming to develop a training concept for technical specialists which uses Virtual Reality and Augmented Reality methods. The project is funded by the German Federal Ministry of Education and Research and started work in February 2020. Its total funding amounts to EUR 1.83 million, and is made up of public funding and contributions from the industrial partners.

Funded by the German Federal Ministry of Education and Research

Further information is available at:

www.mrils.de


PROMISE 4.0 Collaborative Doctoral Programme

At the Faculty of Mechanical Engineering, responsibility for supervising doctoral research on the topic of “Smart Factory Data and Simulation” in the context of Industry 4.0 lies with the Virtual Automation Lab (VAL). The doctoral research is undertaken within the PROMISE 4.0. Collaborative Doctoral Programme 

The project is funded by Baden-Württemberg Ministry for Science, Research and the Arts.

Further information is available at:

Promotionskollegs PROMISE 4.0


Apply for summer semester 2022!

The application period for the summer semester 2022 begins on November 15th 2021 .

More information