The industrial sector is undergoing a profound transformation towards Industry 4.0, driven by the need for increased efficiency and flexibility. A key enabler of this shift is the adoption of wireless communication, which facilitates the use of mobile and flexible systems like autonomous mobile robots (AMRs), while reducing the costs and constraints of traditional wired connections. This has spurred the rise of private cellular networks, a trend now supported by regulatory actions such as the allocation of 400 MHz in the 3.8-4.2 GHz band for local private networks in Europe. However, deploying wireless technology in industrial settings presents significant challenges, including strict requirements for low latency and high reliability, as well as the management of interference and seamless handovers for mobile devices.

This thesis addresses the critical knowledge gap between the theoretical capabilities and real-world performance of industrial wireless technologies. Conducted as part of the 5G-Robot project, a national initiative focused on transforming manufacturing through wireless automation, it presents a comprehensive empirical assessment of different wireless systems, including commercial Wi-Fi 6, MulteFire, 5G Non-Public Networks (NPN), proprietary Industrial Wi-Fi (IPCF-MC), and Ultra-Wideband (UWB). The research systematically analyzes key performance metrics under realistic industrial stressors, including high device load, mobility, and co-channel interference, to provide empirically grounded insights for selecting and deploying fit-for-purpose wireless solutions.

The main contributions of this work are:

• The first experimental results for MulteFire have been presented, establishing a performance benchmark for unlicensed cellular systems in industrial deployments.
• A novel comparative performance analysis of MulteFire, Wi-Fi 6, and 5G NPN (as a proxy for 5G NR-U) to assess their suitability for real-time industrial applications with multiple devices and mobility. This includes a detailed interference evaluation comparing MulteFire and Wi-Fi 6, offering insights into their coexistence challenges in the unlicensed 5 GHz band.
• The first publicly available, real-world latency evaluation of the IPCF-MC, a deterministic industrial Wi-Fi system. The study demonstrates a method to properly configure Quality of Service (QoS) for accurate latency measurements and compares its performance against 5G NPN and Wi-Fi 6.
• A comprehensive empirical investigation of UWB technology, characterizing its performance under various self-interference and propagation conditions for industrial communication links. The study provides practical guidelines for robust UWB system deployment and network planning, emphasizing the importance of frequency separation and modulation choice to mitigate interference.

Our findings reveal a clear performance hierarchy and underscore that the optimal wireless solution is application-dependent, based on a trade-off between peak performance and deterministic reliability. Commercial Wi-Fi 6 offers the lowest median latency in ideal conditions but proves unreliable under mobility and load, exhibiting long latency tails and significant packet loss. MulteFire provides a more reliable alternative for mobile use cases but is constrained by lower throughput. For specialized needs, IPCF-MC successfully delivers a strictly bounded worst-case latency of approximately 128 ms essential for safety-critical control, while UWB is confirmed as a viable solution for sub-millisecond, short-range communication, provided that self-interference is managed through strategic network planning. Ultimately, 5G NPN consistently demonstrates the most robust, scalable, and reliable performance across all scenarios, establishing it as the benchmark for future mission-critical industrial wireless connectivity and validating the ongoing regulatory support for private cellular networks.

After the defence there will be a small reception at Fredrik Bajers Vej 7C/2-209