Hybri-iT Worker

Industrie 4.0

Industrie4.0 will deliver new, digitally refined, intelligent products. Customer will be able to design their own products and have them produced at a reasonable cost. Products and production will be more versatile. Products will be designed by globally distributed design teams and produced in smart factories.

Working in such factories will be way more safer and way more fun. In these “modern times” the work will be more varied. Assistive Systems will support the workers to do their job. Sensors and smart algorithms scan the environment and react to prevent harm. Ergonomic simulation help find the optimal work procedures. Information will be readily available. And collaboration will be easy.


Industrie 4.0 is based on the consolidation of Cyber Physical Systems (CPS) and the Internet of Things and Services. It allows to make simple machines more intelligent and able to communicate information across a network.

Product, Production and Process

Production turns raw materials into products for customers. The production process involves logistics, buildings, energy, workers and the manufacturing of the product itself. These core processes are supported by management processes for planning, control and communication of necessary information. 

The classic production process turning raw materials into products for customers.

In realizing the vision of Industrie 4.0 all parts of the company are augmented through digitalization. It enables workers to gain access to information anytime and anywhere and makes simple devices into Cyber Physical Systems. All these participants than communicate with each other and thus creating an Internet of Things and Services within the company.

Industrie 4.0 vision augmenting each stakeholder in the production process to communicate with each other.

Digital Twin

Industrie4.0 creates a digital twin of the real-life production and product. This has a lot of advantages. First the digital twin yields far easier access to information regarding the state of the production process. Second devices, machines and products can have a broader range of functions and features generating a technological advantage to competitors. Third the digital twin can be used without the real-life counterpart. By replacing the real-life counterpart by simulation enables predictions, allows for training of workers in virtual environments, enables customers to try and test the products up-front on the companies website and to market the products by showing the full range of the functionality as internet based applications. 

Horizontal Integration 

With this crucial information about the production as well as the products can be retained right from the beginning and travel down the value chain up til the customer (horizontal integration). Since communication is way easier for the customer to gain information about the status of his or her product. The company moreover can receive instant feedback from the customer and optimize their processes.

Vertical Integration

Moreover Industrie 4.0 eases the control and adaption of production by making information available to all levels of a company (vertical integration). Bottlenecks could be detected, production could be changed adapting to individual products, setup-times could be reduced and error prone steps eliminated.

4. Industrial Revolution

Industrie4.0 will further efficiency and flexibility in production and will thus reduce cost and increase the number individual products. That is why Industrie4.0 is also called the 4. Industrial Revolution.

The Industrial Revolutions

The 1. Industrial Revolution was the mechanization using water and steam power. With this power large machines could be build and open up the development for mass production.

The 2. Industrial Revolution is marked by the invention of assembly belts and electric energy. While water and steam engines tightly couple energy production and consumption electric energy can be produced anywhere and than very efficiently transported to the places where it is consumed. This allowed for large power plants for producing lots of energy and smaller and this faster production lines like the assembly belt.

The 3. Industrial Revolution was electronics, digitization and information technology. The most remarkable effect was what is called technological convergence. Before the 3. Revolution there were specific devices providing specific services e.g. to listen to music one had to use a phonograph. Nowadays many devices provide these services e.g. televisions, smartphones, computers.

I4.0 Research

The research of the department of Agents and Simulated Reality in the field of Industrie 4.0 focusses on three areas: System Simulation, Integration and Interoperability as well as Assistive Visual Systems.

System Simulation

The research within the topic of Systems Simulation aims at supporting customers, planners and workers to reach their specific goals.

Modern products are designed with the help of computer programs. From small scale things like patient specific implants, over furniture to cars and airplanes. Even large scale structures like cities are visualized in order to gain insights into structures and relationships.

We anticipate how things would look like in the future using Virtual Technologies to depict what could be.

Within ASR we develop System Simulations in a three step process:

  1. Model the reality according to the questions which should be answered.
  2. Run simulations with these models to anticipate the structure or behavior.
  3. Visualize the complex results and make room for human interpretation.


Models are used to describe real-world behavior and structures. In order to fulfill the needs of Industrie4.0 with its complex systems all kinds of models have to be integrated. The models examined within ASR mainly fall into two categories: Models used to simulate the looks and models to simulate behavior. 

The former are geometrical models, material models or lighting models. They are used to render images of complex geometrical structures like cars or airplanes as well as production facilities. But also product models are created. For example the goal of the project IIP-Extrem is to derive 3D-models of patient specific implants. Material models and lighting models are used to predict the looks of products and to map the complex interaction of light bouncing off of different surfaces.

The latter are behavior models of all sorts. That ranges from behavior of humans or robots up to behavior of more complex social systems e.g. pedestrians. This is one of the topics in the research project REACT where pedestrians behavior is used within the application area of autonomous driving. Regarding humans there is also an emphasis on modeling natural human motion. In order to generate synthetic machine learning training data the motion of human pedestrians has to be simulate as correct as possible. Moreover facial expressions also play a crucial role in the communication in road traffic.


Based on the models the goal of ASR is to run interactive or near-interactive simulations in order to predict system structure and behavior. Within Industrie4.0 the simulation of facility layout, individual production steps and worker behavior is crucial to solve the challenges imposed by highly individual products. 

Technologically running interactive simulations of complex models is a compute-intensive task. Our approach to solve this is the employment of High Performance Computing namely through AnyDSL. It allows to implement algorithms in an abstract manner and than automatically map these algorithms to specific hardware platforms while leveraging the full potential of that platform. This drastically reduces the workload for deploying algorithms to different platforms. 

More on the application level are technologies like AJAN and MotSy. AJAN is a multi-agent simulation environment with which high-level behavior can be simulated. This is than e.g. mapped to human motion using MotSy which uses an statistical approach to render variable human motion. 

On an even more abstract level there is the Genesis platform. This platform simulates complex traffic situations including vehicles and pedestrians. The goal is to generate synthetic training data for machine learning algorithms that drive autonomous vehicles.


Modeling complex real-world scenarios and simulating them are only the first two steps. In order to gain understanding one has to come up with sensible visualizations of the simulation results that allow to understand. 

ASR uses a wide range of visualization techniques to reach this goal. From standalone applications, to web-based solutions up to virtual and augmented reality approaches. Each simulation needs its own individual visualization that pays respect to the particular needs of the application scenario.

Integration & Interoperability

Semantic Interoperability

Complex IoT systems deliver a variety of information. They add to the already vast amount of information from product design, reports, specifications, design, production and marketing constraints and so on. Managing this paramount heap of information by hand is difficult if not impossible. Computer based systems are necessary to cope with such a complexity. In order to realize this the information has to be given meaning i.e. raw data has to be turned into semantic data making it readily available to machine processing.

The final goal is to have a company wide semantic world model that incorporates all crucial information about the products, processes and production. This semantic world model allows not only to integrate all necessary information but also to model the relationships between the entities. In Arvida we looked into the question how the data in a virtual planning process can be integrated and the relationships of individual data entities could be modeled. The goal was to make legacy systems more interoperable.

The principles of Linked Data were also applied in the project Archimedes where we looked into how  heterogeneous information about aircraft cabins could be combined into a single world view to make well founded decisions about cabin design.


Turning machines and devices into Cyber Physical Systems enables them to communicate and take part in the internet of things and services. But this also poses security risks. Within ASR we also look into the security aspects of Industrie4.0.

Secure communication is key to a successful realization of the Industrie4.0 vision. Within the project Designetz we develop solutions how devices can be made secure while still being able to unfold their fill cyber physical potential. 

ASR also provides the IoT-Testing Laboratory where manufacturers can certify their IoT products. The lab is part of the Test Center for IT-Security (PITS) which provides common criteria certification services and is one of the few german test centers. 

Assistive Visual Systems

Assistive Visual Systems in the sense of the research within ASR yield interactive visual support within Industrie4.0. Building on individual system’s simulation which are interoperable Assistive Visual Systems integrated these building blocks into large scale applications that span multiple domains vertically as well as horizontally. Thus enabling a holistic view of all the involved process steps and allowing overarching optimizations.

Within ASR this global vision comes with two main areas of application. First Collaborative Design and Decision Making and second Condition Monitoring & Smart Predictive Maintenance . 

Collaborative Design and Decision Making

Todays product design is mainly done by digital tools involving experts from different domains. Making collaboration between these experts possible one has to bridge the boundaries of the tools that these experts use. In the project Arvida ASR developed ways to make these boundaries more permeable. 

The review application developed for BMW we created a web based application that enabled experts to collaborate in a joint review process to detect errors and make proposals for product enhancements. The project Collaborate3D funded by the BMBF took these ideas one step further and augmented design models of products by semantic information aggregating information from different domains into a holistic view of the product. The validity of this approach was proven by successfully transferring these approaches to whole different domain with the project virtruv21 which tackled questions regarding international construction projects.

But these techniques not only make collaboration between experts more easy. Having virtual models of products available also customers can also use these to configure the products to their wishes and needs. ASR has looked into this aspect in different collaborations with industry partners. Adding the possibility to use the design models also for marketing purposes helps amortize the design costs. One example is the project of the Airbus Cabin Configurator. 

Highly individual products also impose a rising amount of flexibility on the production. Facilities have to be planned with respect to these needs, machine setup times be reduced. Immersive planning environments support production planners reducing the costs. In the EU-funded projects FITMAN ASR looked into this aspect. 

These virtual environments can moreover be used to simulate the workflows of workers in these environments. Not only to optimize the work but also to look into ergonomic aspects of the individual steps of the work. In the project Interact we investigated questions regarding ergonomic questions in the area of a automotive assembly process together with industry partners.

These technologies can further be used to train workers and thus make the transition from creating one variant of the product to the other more easy. Interactive support during production helps the worker cope with the different requirements. Team collaboration is also enhanced. With the project Hybri-iT we look into questions that arise when human workers and robots closely collaborate on common tasks.

This opens the doors for a holistic process planning and control. Dynamic planning and suggestions for production planners can be given. AI methods can be employed to optimize process planning. These semantic end-to-end planning and suggestions where examined within the scope of several projects of the team Intelligent Informations Systems.

Condition Monitoring & Smart Predictive Maintenance

Condition Monitoring & Smart Predictive Maintenance the state of individual systems is monitored and prediction of future states of the systems can be made. Integrating the data coming from different IoT-Systems can quickly lead to information overload. 

The first step is data aggregation in which all relevant information is gathered. This in itself is a challenge given the large number of IoT data sources with different protocols, granularity, frequencies and formats. Intelligent dashboards filter and map relevant information and make them readily available to decision makers. They combine the different systems into a digital twin of the physical systems and thus create the dual reality of physical and virtual world inherent to Industrie4.0. 

On a technological level ASR relies on semantic technologies to make simple data e.g. the number 21 into smart data by augmenting the data with meaning and turning the number 21 into useful information like 21° Celsius on 27. Sept. 2018, 12:11:45 at metering point #45 of oil pump #72. 

IoT generate a vast amount of sensor data. Semantic Stream Reasoning enables Industrie4.0 to filter this data and look for patterns and abnorm system behavior thus yielding symptoms of possible faults. This ranges from simple out-of-range cases e.g. temperature to high up til more complex cases combing smart data from different metering points. 

As in the medical case the symptoms are diagnosed in order to detect possible faults. From these faults the overall system condition can be derived using a belief network. ASR has employed these technologies in the two industry cooperations with Hydac ICM Wind and ICM Hybraulik. 

But from the symptoms not only the current condition can be derived. Using hybrid automata the future faults can be predicted. Based on the current state of the symptoms using a hybrid model of the production process possible fault-states can be derived. This fault prediction was used in the project INVERSIV to predict future fail conditions of a production process.

These technologies lead to fault reduction or fault prevention. Downtimes are reduced and productivity is increased. Moreover the quality of the process and the products is enhanced. Last but surely not the least the safety at work can be raised if possible dangerous situations can be detected prior to machine failure.


Industrie4.0 promises to move todays industry to the next level. Not only new products are expected. But also production processes will dramatically change making work more interesting and safe than it is today. 

Within ASR we do research into simulating individual systems to make predictions, integrate them into more complex systems yielding Assistive Visual Systems that support workers, planners and customers.




  • Fastlane - Nowadays Web applications, including 3D virtual worlds, real-time simulations, and virtual reality applications demand a high amount of processing power.… Continue reading Fastlane
  • AJAN - AJAN (Access Java Agent Nucleus) is a modular agent web service which integrates different AI technologies in an intuitive way… Continue reading AJAN
  • AnyDSL - AnyDSL - A Framework for Rapid Development of Domain-Specific Libraries AnyDSL is a framework for domain-specific libraries (DSLs). These are… Continue reading AnyDSL
  • MotionGraph - Morphable Graph [1] is a generative, graph-based approach for data-driven motion modeling and synthesis. Motion capture data is represented by… Continue reading MotionGraph
  • I2S Software - For more information on our open-source software (tools, benchmarks, system demos) for semantic service coordination and data analysis, please visit:… Continue reading I2S Software
  • ATLAS - ATLAS (which stands for "Advanced Three-dimensional Large-scale Asset Server") is a system for storage and conversion of polygonal 3D models.… Continue reading ATLAS
  • COMPASS - COMPASS ("Collaborative Modular Prototyping And Simulation Server") is a framework for developing web applications with which elements of 3D scenes… Continue reading COMPASS