The further development of production systems through information and communication technologies into a network of services, resources, robots and products enables adaptive process chains and short response times to disruptions and

Fluctuations. Individual processes and resources are no longer considered locally, but interact dynamically with each other across the board. This creates the potential to break up rigid process chains and create dynamic and linear assembly systems, especially in assembly, which is the last link in the value chain and is particularly affected by fluctuations.

As robots in assembly in particular were previously seen as closed systems, they and their processes are encapsulated. Conventional robot controllers offer little scope for more complex and computationally intensive applications beyond their basic functions. Examples of this are machine learning algorithms or a connection to higher-level networks for evaluating large amounts of data (big data).

Edge and cloud networked

One solution approach is an edge cloud architecture. Applications such as web robot programming, dashboards and databases are available factory-wide at the cloud level. The factory cloud can be hosted in the same network or in outsourced networks. A connection to the edge level can be established via common non-time-critical networks. Process data can be aggregated and transferred from the edge to the cloud. Applications packed as containers are rolled out from the cloud to the edge server and executed there. This roll-out can be structured using DevOps and automated using continuous deployment/continuous integration. These applications can be robot path planning, algorithms for image processing or applications for the localization and navigation of mobile robots, for example.

The applications communicate directly with the store floor level. This is only possible if real-time requirements for control and regulation such as low latency, high reliability and high availability are met by both the communication infrastructure and the edge. The Time Sensitive Networks Standard (TSN) can serve as the basis for deterministic Ethernet networks that enable low latency and high availability. The 5G mobile communications standard also enables wireless real-time communication in the "Ultra Reliable and Low Latency Communications (uRLLC)" configuration. On the hardware side, IIoT gateways can be used at store floor level as devices to enable real-time communication between robot controllers or other field devices. This capability is implemented through the use of drivers, wireless modules and protocols.

Autonomous and adaptable robots

The vertical networking of robots and the outsourcing of computing power open up new fields of application that are not possible with local control systems. One example is an autonomous mobile robot (AMR) with fast component recognition in conjunction with adaptive path planning to quickly adapt processes to new products. In addition to real-time wireless communication, the high computing power available on the edge and the edge's ability to quickly process image data as required using graphical processing units (GPUs) are utilized for this purpose.

With the exception of safety systems and controllers, which have to access the robot locally due to a possible network failure, outsourcing the computing power would make it possible to operate robots without cost-intensive dedicated robot controllers. Another example of a new field of application is the programming of robots from different manufacturers with a uniform graphical user interface (GUI) or a standardized programming language. The "Robot Operating System" middleware suite, which offers many robot models and drivers as well as algorithms for path planning and localization, is suitable for this purpose. The challenges in implementing an edge cloud architecture for robots are the integration of a real-time capable, reliable network in a brown field environment, the standardization of communication protocols and a rethink towards cloud and DevOps-based software development. Amon Göppert

Briefly explained: The MHI e.V.

The Wissenschaftliche Gesellschaft für Montage, Handhabung und Industrierobotik e.V. (MHI e.V.) is a network of renowned university professors - institute directors and chair holders - from German-speaking countries. The members conduct both fundamental and application-oriented research on a wide range of current topics in the fields of assembly, handling and industrial robotics. Further information on the society, its members and activities: http://www.wgmhi.de

Briefly explained: The WZL at RWTH Aachen University

The Machine Tool Laboratory WZL at RWTH Aachen University is headed by the four professors Thomas Bergs (Manufacturing Processes, Gear Technology), Christian Brecher (Machine Tools, Cluster of Excellence, Gear Technology), Robert Schmitt (Production Metrology, Quality Management, Assembly Technology) and Günther Schuh (Production Systems). The WZL conducts research into current topics in production technology and has extensive previous experience in areas such as robotics, sensor technology, real-time communication and assembly. The research carried out at the WZL is characterized by close cooperation between the various disciplines. As a rule, research work is carried out together with a wide range of industrial companies. This ensures that the research results are quickly transferred into operational practice and that the research developments address a need in industrial application.