Running time-critical and computationally intensive processes of mobile robots on edge computing infrastructures is a key enabler of flexibility, scalability, and decentralized control. The real-time performance of such off-loaded control tasks strongly depends on the quality and stability of the underlying wireless communication link. In environments with a coexistence of multiple wireless access technologies, this becomes even more challenging. The present paper sets forth an experimental system that combines a nonlinear model predictive control (NMPC) strategy with dynamic switching of wireless connectivity between 5G, WiFi 5 GHz and WiFi 6 GHz. The control logic is executed on an edge server and transmitted to an Automated Guided Vehicle (AGV) over a set of different wireless channels. The proposed architecture is validated through real-time trajectory tracking in a laboratory environment. The findings of this study demonstrate that robust switching mechanisms and network awareness are crucial for stable operation under realistic wireless conditions, underscoring the importance of co-designing communication and control systems for edge-cloud-based robotic applications.

In this paper, we propose a multispectral deep fusion network with distance estimation to ensure the safety of persons. Autonomous mobile robots (AMRs) have to detect persons in all weather and lighting conditions. By fusing sensors in the thermal and visible spectrum, fewer persons are missed. We introduce a detail module, which dynamically adjusts the weights of the modalities depending on the information density of the RGB and thermal images. We evaluate the proposed multispectral fusion method on the FLIR dataset. If the distance to the detected person is unknown, the AMR cannot judge whether it needs to brake or adjust its path. We therefore propose a distance estimation, which we evaluate on the KITTI dataset. To evaluate the multispectral fusion method as well as the distance estimation, we recorded and labelled a dataset from the perspective of an AMR in the visible and thermal spectrum with distance information for each object. On this dataset, we evaluate the performance of our network and discuss the number of dangerous failures for functional safety applications, and thereby, contribute to safe outdoor applications.

Wireless networks have made great strides in recent years in supporting the Industrial Internet of Things (IIoT). However, industrial network security remains a major challenge in the context of wireless access. Notably, many industrial communication services that enable factory floor applications are still based on proprietary and industry-specific protocols that lack essential security features. As a solution to this challenge, in this paper, we propose an application function (AF) based security framework (AERO). This framework enables the 5G security function to meet the security requirements of industrial networks when wireless access is used without implementing redundant mechanisms. The basic idea is that the 5G system implements the user plane cryptographic mechanism over the air, based on industry requirements, without affecting existing industry protocols. We propose ($i$) static configuration and (ii) dynamic configuration mechanisms that facilitate providing the 5G core network security policy used to determine user plane cryptographic requirements. Furthermore, we demonstrate the advantages of our AERO approach by applying it to two specific use cases: Automated Guided Vehicles (AGVs) and Electrified Mono-rail Systems (EMSs). Through our analysis, we establish that our proposal significantly reduces data transmission time when compared to the current 5G mechanism.

To enhance productivity, assist workers and to have a more flexible material flow, autonomous mobile robots (AMRs) are used in industrial applications. To guarantee the safety of workers, most AMRs are equipped with safety laser scanners and stop if a person enters the protective field. To leverage the benefits of AMRs outdoors and in urban logistics, several new challenges arise. Severe weather conditions like rain, fog, snow, low lighting and direct sunlight restrict the capabilities of cameras in the visible spectrum, thermal cameras, laser scanners and time of flight cameras. Besides the safe detection of persons, the distance from the robot to the person has to be measured or estimated to provide a classification if the person intrudes the protective field of the AMR. The state of the art pedestrian detectors are based on neural networks and designed for advanced driver assistance systems. Therefore, most datasets for training the networks are acquired from the rooftop of cars and do not reflect the perspective and environment of AMRs. Our study reviews the state of the art sensors and pedestrian detection methods and shows the benefit of sensor fusion technologies based on artificial intelligence. Further, we discuss the limitations of these methods for industrial outdoor and urban AMR safety applications and propose methods to overcome these drawbacks.

With 5G technology and its application in industry, the demand for wireless communication is increasing and so is the need for information about the wireless channel. Thus, accurate propagation channel models are required. For this purpose, we analyze our recently measured and published Power Delay Profiles (PDPs) in industrial scenarios using the Saleh Valenzuela (SV) model, compare the results with the results of other scenarios and propose an improved model. Further, we discuss the relation of path loss models, PDPs and the SV-model.

Due to an increasing need for mobility on the factory floor, future industrial applications will often use wireless technologies like 5G. The necessarily high reliability not only needs to be designed into the wireless communication technology, but also has to be experimentally verified. Events that degrade the reliability of highly available systems are rare. Additionally, correlation is inherent to wireless channels and an analysis method that is able to cope with those two challenges is needed. Thus, in this work our aim is to provide such a method. We evaluate the practical usage of the limited relative error (LRE) algorithm and discuss how to obtain confidence intervals based on it. Thereby, the standard normal distribution is approximated, which leads to biased results. We illustrate this by the usage of the Berry-Esseen theorem. To overcome the bias the LRE algorithm is enhanced with moving block bootstrapping (MBB). We call this new method bootstrapping LRE algorithm. We use measurement data from a 5G network for our analysis. Further, the usability of the 5G network for communication purposes based on the black channel paradigm is evaluated with the 5G measurement data and its statistical analysis. Thereby, the performance parameters of cycle time and reliability of the black channel in 5G are determined.

Bandwidth is a scarce resource in industrial wireless networks. At the same time keep-alive-messages are excessively used in these networks to test channel characteristics. This work proposes an approach to reduce bandwidth usage in critical industrial applications, by exploiting channel quality data from 5G. Thereby, the information about the channel is gained from channel state information (CSI) and the radio resource control (RRC) state. This work combines the CSI with the RRC state to determine whether a wireless node is connected. Subsequently, the connection information is exposed towards the application server via the network exposure function (NEF) of the 5G core network and to the mobile application by the UE. Since the CSI and RRC state are always present, this results in a net bandwidth reduction on the air interface of approximately 1.65 Mbps in the examined reference use case in contrast to the standard keep-alive mechanism.

In industrial applications continuous wireless connectivity of mobile clients can rarely by guaranteed. Lack of communication negatively impacts the performance of industrial automation systems, e.g. Automated Guided Vehicle (AGV) fleets. Utilizing industrial Mobile Ad-hoc NETworks (MANETs) and adaptive positioning systems can reduce the number of disconnections in these AGV fleets. Therefore the performance of the mobile systems (e.g. AGV fleet) is improved and factory efficiency increased.

Industry 4.0 envisions flexibility and mobility to be highly important for the factory of the future. Mobile wireless networks are important tools, that enable flexibility and mobility. A better understanding of the mobility of industrial nodes is necessary for the design and implementation of effective mobile wireless networking solutions.In this work, the mobility of industrial clients is characterized. The position of 30 Automated Guided Vehicles (AGVs) is recorded and analyzed. The applicability of common models like RWPM and the Manhattan model are compared to the empirical data and improved by a newly proposed algorithm for the selection of node speeds. A new metric for the characterization of the impact of the mobility model on the type of propagation path in a wireless network is introduced.

Wireless communication continuously gains in importance in the context of industrial automation due to its role as an enabling factor for flexibility. The number of mobile, wirelessly communicating devices in modern production facilities steadily rises. Knowledge about the signal propagation is central to the planning and maintenance of these wireless networks. Precise models enable users to supply sufficient communication capabilities to the factory of the future.Measurements of Power Delay Profiles (PDPs) were performed at the frequencies of 2.1,2.6,3.8 and 5.1GHz for three different scenarios in a modern and running production facility. The results are analyzed by means of multi-path components, delay spread and path loss. The results are compared to previous measurements. Thereby, further insights in propagation conditions in industrial scenarios are gained.

Tomorrow's production facilities will be more flexible and efficient by incorporating new technologies, like massive wireless sensor networks, intelligent logistics solutions, human-machine-cooperation and a tight-knit communication connecting all of these systems. The rise of mobile machines and the required ability for reconfiguration of static production facilities fosters the use of wireless communication technologies. However even communication technologies that provide features specialized for industry, like 5G, will not be able to fulfill all requirements of the industrial environment simultaneously. We show this by analyzing the requirements of industrial use cases and comparing them to the performance of current and future communication technologies. We introduce current fields of research, which focus on technologies and strategies used in the factory context for solving these challenges. To do so the factory of the future will use multiple communication technologies and those need to be orchestrated properly.