dtec.bw Jahrestagungen

Drahtloses Sensor Netzwerk zur Form- und Dehnungsmessung für Großstrukturen

Motivation

Die Echtzeit-Formvermessung großer Bauteile ist in industriellen Fertigungsumgebungen bislang nur eingeschränkt möglich. Das entwickelte Sensornetzwerk erfasst die Bauteiloberfläche drahtlos und parallel mit mehreren Messknoten. Dadurch bleibt die Messzeit weitgehend unabhängig von der Bauteilgröße und das System lässt sich flexibel skalieren. So entsteht die Grundlage für eine kontinuierliche Prozessüberwachung und Hardware-in-the-Loop-Fertigungssysteme, sowie Zeiteinsparungen in den Fertigungsprozessen.

DoS Robustheit für Embedded Geräte und Essentielle Funktionen

Motivation

Denial-of-Service (DoS) Angriffe werden hauptsächlich mit Angriffszielen und Technologien aus der Internet-Domäne in Verbindung gebracht. Allerdings sind auch Embedded Geräte wie industrielle Steuerungen von DoS Angriffen beeinflussbar. Angriffe auf kritische Infrastruktur der Ukraine zeigen auf, dass dies reale Angriffe sind welche aktiv eingesetzt werden. Im Rahmen des Projektes DS2CCP ist es Ziel, bewährte Verfahren und Lösungen für Embedded Geräte aufzustellen um DoS Robustheit zu erhöhen.

Source-Code-basierte Generierung einer Software Bill of Materials

Motivation

In der Vergangenheit wurden Schwachstellen in der Software Supply Chain ausgenutzt, wie z.B. durch den xz-Angriff. Um die Behebung dieser Schwachstellen zu unterstützen ist eine Software Bill of Materials (SBOM) geeignet. Eine SBOM bezeichnet eine Stückliste, welche die Komponenten einer Software aufzeigt. Verbreitete Standards sind: CycloneDX und SPDX. Die Generierung erfolgt im Regelfall mit SBOM Tools aus Dateien, welche der Paketmanager zur Verfügung stellt, aber unter Programmiersprachen wie C ist die SBOM Tool Unterstützung gering und der Paketmanager Conan nicht verbreitet.

Safety and Security: I0-Link Wireless and OPC UA over 5G under prEN 50742

Motivation

Challenge: Emerging regulations like prEN 50742 mandate the integration of cryptographic security mechanisms directly into safety-critical data streams.
Problem: While cryptography enhances security, its timing and packet size overheads might violate safety watchdogs and reduce system availability.
Objective: Empirical evaluation of safety-security convergence across a complete control chain, from an IO-Link Wireless Safety (IOLWS) device to a PLC via an OPC UA backbone.

Drone Detection over 5G Cellular Networks

Motivation

Drones are increasingly used for surveillance, delivery, inspection, and autonomous operations, but unauthorized drones can also create serious safety and security threats, including intrusion into restricted areas, disruption of airports, illegal surveillance, smuggling, and potential attacks on critical infrastructure.
Modern drones can use public or private 5G networks for command and control, telemetry, navigation data, and highbandwidth video streaming, making cellular connectivity an important part of their operational footprint.
Conventional RF-based drone detection mainly observes over-the-air emissions and may struggle with interference, encrypted communication, or drones using standard cellular links.
Our goal is to detect and identify probable drone activity from 5G network observations by correlating RAN behaviour, control-plane signalling, UE/session information, and userplane traffic patterns.

Interferometry for Sensor Networks

Motivation

Interferometry provides: Ultra-high precision measurement, Dynamic sensing capability
Sensor network offers: Scalable sensor nodes with the demands, Providing adaptive responses, Yielding insights for operational efficiency
Our Objectives?
– Compact, cost-effective, high-precision optical sensor head: incorporation of the technique, „Deep frequency modulation interferometry“ to reducing optical complexity, with low-cost commercial off-the-shelf lasers
– Integration of optical sensing with distributed networks
– Extending sensing range via test mass design aspects such as: retroreflector (corner-cube), plane mirror or converging optics (concave mirror)
– Preferably wireless, low-power sensor node
From industrial applications to the research facilities, these kinds of measurements are crucial for operation.

Security Monitoring of OPC UA over Industrial Private 5G Networks

Motivation

Industrial private Fifth Generation (5G) networks enable flexible wireless communication for industrial applications, but realistic cybersecurity evaluation requires representative experimental environments and monitoring infrastructures. Building on such an environment, machine learning (ML)-based intrusion detection systems (IDSs) have demonstrated that attacks against Open Platform Communications Unified Architecture (OPC UA) over industrial private 5G can be detected using traffic- and protocol-aware features. When OPC UA uses SignAndEncrypt, payload inspection is no longer possible. Intrusion detection must therefore rely on residual transport, temporal, and protocol-lifecycle characteristics that remain observable in encrypted traffic. However, benign 5G connectivity variations such as user equipment (UE) reconnections, Protocol Data Unit (PDU) session resets, and temporary interruptions can noticeably increase the falsepositive rate (FPR), even in the absence of attacks. These findings motivate future 5G control-plane (CP)-aware intrusion detection to improve IDS reliability under dynamic network conditions.

Assessing the Controllability of Modular UAV Swarms over 5G Networks via ROS2

Abstract

Multi-UAV operations are constrained by legacy point-to-point radio links that scale poorly in range, bandwidth management, and quality of service. In this feasibility study, the controllability and reliability of a highly modular ROS 2 based UAV swarm interconnected via private 5G cellular networks are evaluated. After integrating multiple aerial vehicles into a 5G architecture, the communication performance is compared against standard WLAN by profiling end-to-end latency and throughput under representative mission and sensor traffic. The findings provide practical design guidelines for deploying UAV swarms over cellular infrastructure, particularly in Beyond-Line-of-Sight (BLoS) scenarios, where scalability and coverage outweigh raw latency performance.

Adaptive C-UAS Swarm with Ad-hoc 5G-SA Network

Abstract

Threat scenarios with hostile Unmanned Aerial Vehicles (UAVs) are becoming increasingly difficult to handle. Additional to the variety of UAV types with their different capabilities, the situation will become even more difficult when Global Navigation Satellite Systems (GNSS) are occasionally jammed or unavailable. Therefore, C-UAS (Counter-Unmanned Aerial Systems) need to have a robust communication for a reliable operation. Our C-UAS solution, where multiple UAVs interact to prevent a potential threat, operates agile and autonomously to cover a broad range of scenarios also in hostile environments. By using an ad-hoc 5G airborne network with AI-based carrier frequency selection, swarm communication in jammed environments will be feasible. Communication traffic will be exchanged between Public 5G Mobile Networks and the Private HSU 5G Campus Network, established under the dtec.bw-DS2CCP project.

Security in Private 5G Campus Networks

Motivation

Die moderne Fertigung ist durch einen hohen Automatisierungsgrad geprägt, wobei insbesondere im Bereich funktionaler Sicherheitsanwendungen potenzielle Gefährdungen für Menschen und Anlagen strikt zu vermeiden sind. Private 5G-Campusnetze bieten in diesem Kontext eine hohe Flexibilität und Modularität sowie einen reduzierten Aufwand für Installation und Wartung. In zeitgemäßen Kommunikationssystemen hat die Gewährleistung der Cybersicherheit – wie im Cyber Resilience Act (CRA) betont – oberste Priorität.Im Rahmen des Projekts Digital Sensor-2-Cloud Campus Platform (DS2CCP) wird daher eine auf Künstlicher Intelligenz (KI) basierende Lösung zur Deep Packet Inspection (DPI) in die private 5G-Umgebung integriert. Ziel dieser Lösung ist unter anderem die Erkennung potenzieller Anomalien in der drahtlosen Kommunikation zwischen der industriellen Produktionsebene und der Edge-Cloud.Durch diese Integration soll eine sichere und funktional zuverlässige Testumgebung für Sicherheitsanwendungen geschaffen werden, die innerhalb des Campusnetzes kontinuierlich überwacht und bewertet werden kann. Die KI-gestützte DPI-Prüflösung stellt dabei einen zentralen Bestandteil einer umfassenden Sicherheitsstrategie dar, welche die digitale Souveränität innerhalb des Netzwerks gewährleistet und zugleich einen digitalen Zwilling des Kommunikationsverkehrs abbildet.

DoS Robustheit für Embedded Geräte und Essentielle Funktionen

Motivation

Denial-of-Service (DoS) Angriffe werden hauptsächlich mit Angriffszielen und Technologien aus der Internet-Domäne in Verbindung gebracht. Allerdings sind auch Embedded Geräte wie zum Beispiel industrielle Steuerungen von DoS Angriffen beeinflussbar [1]. Angriffe auf kritische Infrastruktur der Ukraine zeigen auf, dass dies reale Angriffe sind welche aktiv eingesetzt werden [2]. Derzeit gibt es keine auf Embedded Geräte bezogenen Best Practices um diese gegen DoS Angriffe möglichst robust zu entwerfen. Stattdessen werden hier Praktiken aus der Internet- und Netzwerkinfrastrukturdomäne herangezogen. Im Rahmen des Projektes DS2CCP ist es Ziel, bewährte Verfahren und Lösungen für Embedded Geräte aufzustellen um DoS Robustheit zu erhöhen.

VR-Enhanced Wireless Safety E-Stop System

Motivation

Traditionelle kabelgebundene Not-Aus-Systeme sind in Flexibilität, Skalierbarkeit und Einsatzfähigkeit in dynamischen Umgebungen begrenzt
Integration modernster Technologien: IO-Link Wireless Safety (IOLWS), OPC UA Safety und 5G-Kommunikation für zuverlässige drahtlose Sicherheitslösungen
VR-basierte Visualisierungsplattform ermöglicht immersive Demonstrationen und fördert operatives Verständnis durch Echtzeitüberwachung

Die Lieferkette als kritischer Faktor für Software Sicherheit

Motivation

Die Sicherheit der Software Lieferkette als ist ein kritischer Faktor. Sowohl der Cyber Resilience Act als auch die Presidential Order aus dem Jahr 2021 fordern eine sogenannte Software Bill of Materials (SBOM), welche die Komponenten und Abhängigkeiten einer Software aufweist. Verschiedene Stakeholder profitieren davon, wie Anwender, Käufer, oder Entwickler.
Anwender profitieren von einer schnellen Behebung von Schwachstellen in Software, wenn ist ihre SBOM aktuell ist. Käufern ist es möglich durch eine SBOM bereits vor dem Kauf sich einen Überblick über die Software zu verschaffen. Komponenten, welche unerwünscht sind, können daher bereits präventiv ausgeschlossen werden. Entwickler können sich bereits während des Entstehungsprozesses einen Überblick über die Software verschaffen, um ungenutzte Abhängigkeiten zu vermeiden und sicherheitskritische Fehler in der Software zu beheben.

Cyber Physical Finite Element Sensor Network (CPFEN)

Motivation

Vermessung von Großstrukturen in Echt-Zeit
Verwendung eines drahtlosen Sensor-Netzwerkes
Erfassung von Form- und Lageänderungen
Messung der mech. Zug-, Druck und Schubspannungen im Material

Roaming Wireless Safety Emergency Stop

Motivation

Modern manufacturing relies on a high degree of automation, where human intervention is kept minimal but is crucial during malfunctions or maintenance. In such an environment the need for secure, reliable, fast and flexible wireless communication solutions is ubiquitous. This research is part of the „Digital Sensor-2-Cloud Campus Platform“ (DS2CCP) project [1], which aims to demonstrate reliable wireless communication between the industrial shop floor and the edge cloud. The goal is to provide a portable emergency stop that operates safely across multiple automation cells. Therefore, the system integrates IO-Link Wireless (IOLW) with IO-Link Safety with a  focus on roaming features [3] to IOLW Safety (IOLWS).

Security in Private 5G Campus Network

Abstract

Modern manufacturing relies on a high degree of automation, where especially for functional safety applications hazards for humans and equipment must be prevented. Private 5G campus networks offer high flexibility, modularity, reduced installation and maintenance efforts. In modern communication systems, cybersecurity is of paramount importance, as emphasized by the Cyber Resilience Act (CRA). Therefore, a probing solution utilizing an AI-based deep packet inspection (DPI) is integrated within the private 5G environment, which is part of the „Digital Sensor-2-Cloud Campus Platform“ (DS2CCP) [1] project, aiming to detect, e.g. potential anomalies in wireless communication between the industrial shop floor and the edge cloud. The goal is to provide a secure and functional safe test environment for safety applications being monitored and evaluated within the campus network. The AI-based deep packet inspection probing solution is part of a security strategy offering sovereignty within the network itself and deploying a digital twin of the communication traffic.

Robustness Testing for Embedded Devices Against DoS Attacks

Motivation

Denial-of-service (DoS) attacks have garnered significant attention in both industry and research for decades due to their capacity to inflict damage using relatively simplistic methods and minimal expertise. However, the topic is mainly discussed in relation to the Internet and network level technologies as well as use cases. It is important to note that Industrial Control System components, such as Programmable Logic Controller and other real-time devices, are susceptible to DoS attacks as well. This is demonstrated in research scenarios like the malware PLC-Blaster [1] and in practical instances such as Industroyer [2].

Currently, no best practice against DoS attacks on embedded devices seem to exist. This research is part of the „Digital Sensor-2-Cloud Campus Platform“ (DS2CCP) project [3], which aims to demonstrate reliable communication between the industrial shop floor and the edge cloud. The main goal is to provide a set of best practice methods as well as solutions to increase the resilience of embedded devices against network based DoS attacks.

Accuracy Evaluation of SBOM Tools for Web Applications and System-Level Software

Motivation

Recent vulnerabilities in software like Log4j [1] raise the question whether the software supply chain is secured sufficiently. Governmental initiatives in the United States (US) [3] and the European Union (EU) [3] demand a Software Bill of Materials (SBOM) for solving this issue. An SBOM has to be produced by using creation tools and it has to be accurate and complete. In the past, there had been investigations in this field of research. However, no detailed investigation of several tools producing SBOMs has been conducted regarding accuracy and reliability. For this reason, we present a selection of four popular programming languages: Python, C, Rust and Typescript. For web application software we consider Python and Typescript while for system-level software C and Rust are investigated.

Security Considerations for IEEE 802.1 Time-Sensitive Networking in Converged Industrial Networks

Abstract

Cyber security becomes more and more relevant for the domain of Industrial Control System (ICS). An aspect, which increases the attack surface of those devices is the trend of Industry 4.0 and the associated network interconnections. While those devices were air-gapped and communication was clearly segregated, new technologies arise, which break up with this concept, since horizontal and vertical interconnection is essential for future use cases in ICS. Time Sensitive Networking (TSN) represents such a new technology, which allows the transmission of hard real-time traffic, commonly present within the field level communication, converged with other communication streams, e.g. non-time critical best-effort traffic, on the same wire. On the one hand this approach brings many benefits for ICS environments, e.g. predictive maintenance to reduce unplanned downtime, logging and others. But on the other hand this technology enlarges the attack surface of ICS and must therefore be analyzed from a cyber security perspective. For instance the transmission of real-time traffic can be disturbed easily by various network-based attacks, which makes protective measures necessary. Theretore, within this work, security considerations for the use of TSN within ICS applications are presented, suitable protective measures as well as potential enhancements are depicted. The threats and mitigations presented within this work are intended to draw attention towards cyber security within TSN based converged networks and provide an overview for possible protection strategies.

Cyber Physical Finite Element Sensor Network

Motivation

In modern manufacturing processes it is mandatory that the components to be manufactured are produced within well-defined tolerances. If only small production tolerances are permitted this often means that all production steps have to be monitored, which is a challenging task, e.g. in the production of large-scale components, where it is often not possible to determine the exact shape or surface stress in real-time during forming or deformation processes. Here, a standard solution is the use of laser trackers, typically requiring a human operator or a robot during the measurement process. Also, measurement time increases drastically with the size of the component. In this project we suggest an approach, where we measure the shape of an object with a finite grid of wireless sensor probes in real-time during the fabrication process. The sensor network measures the surface at discrete points in a similar way classical finite elements (FEM) discretize a volume for modeling the physical behavior. Hence, this measurement system is called a Cyber-Physical Finite-Element-Network (CPFEN).

Interferometry for Wireless Networks

Introduction

Adopting scalable compact wireless sensor network in manufacturing enables the flexibilization in industrial production processes by providing precise, real-time and adaptive responses to structural changes, e.g., the deformation of a curved surface, as part of the dtec.bw project, Digital Sensor-2-Cloud Campus Platform (DS2CCP), with the mutual-reference to the work “Cyber Physical Finite Element Sensor Network” by L. -M. Bretthauer et al.

Adaptive C-UAS Swarm with Ad-hoc 5G-SA Networks

Abstract

Threat scenarios with hostile Unmanned Aerial Vehicles (UAVs) are becoming increasingly difficult to handle. Additional to the variety of UAV types with their different capabilities, the situation will become even more difficult when Global Navigation Satellite Systems (GNSS) are occasionally jammed or unavailable [1]. Therefore, C-UAS (Counter-Unmanned Aerial Systems) need to have a robust communication for a reliable operation. Our C-UAS solution, where multiple UAVs interact to prevent a potential threat, operates agile and autonomously to cover a broad range of scenarios also in hostile environments. By using an ad-hoc 5G airborne network with AI-based carrier selection, swarm communication in jammed environmefrequencynts will be feasible. Communication traffic will be exchanged between Public 5G Mobile Networks and the Private HSU 5G Campus Network, established under the dtec.bw-DS2CCP project [2].

UAV Detection with Color and Infrared Images in the 5G Network

Project Overview

• Detection of unmanned aerial vehicles (UAVs) on color and infrared camera images through machine learning
• Transmission of image and control data between camera and processing computer via the 5G campus network