5G-Based Real-Time Location System in Automotive Production

Over three years, Christoph Küpper investigated different localization technologies and built test fields with various mobile network and hardware providers, including a 5G-based localization setup. With his doctoral research, Küpper broke new ground in largely unexplored technological terrain.

This article presents BMW Group’s vendor-independent localization platform IPS-i, briefly compares the performance of UWB and Wi-Fi, and outlines three 5G test fields and use cases. The dissertation concludes with a surprising result: 5G can deliver high-performance localization – but it still has its limits. At least today.

Küpper is an expert in industrial radio applications and has worked at BMW Group for several years. Since February 2023, he has been Product Owner Industrial Radio Applications, responsible for standardizing and optimizing radio systems in industrial production. In 2024, he completed his PhD on 5G-based localization (RTLS) and analyzed its potential for industrial applications.

IPS-i (International Production System – Identification & Localization) is a localization platform developed by BMW and continuously expanded since 2016 to provide location data as a platform and service solution. Many localization solutions, such as UWB from various vendors, are proprietary and tied to specific hardware.

BMW deliberately chose an open, modular platform approach. The vision of IPS-i is to localize all objects involved in the production process, independent of the technology used. The platform provides position data for diverse, independent services and business processes – scalable and worldwide.

IPS-i integrates technologies such as UWB, GPS, and RFID, which deliver standardized localization data to the platform. Thanks to its vendor-independent architecture, BMW plants can flexibly select the technology best suited to their needs without being tied to a specific supplier.

The platform collects and processes position data from various sources, consolidating them into a unified database. This enables technology-independent analysis and simplifies integration into existing IT infrastructures. Data aggregation and normalization standardize position data from different systems.

A scalable application processes over 500 million data points per hour across 14 plants worldwide.

The service layer allows applications and systems to access standardized position information directly, without dealing with the details of the underlying positioning technology. Services and use cases can be configured in a self-service model by local users in the production area.

UWB is one of the most precise localization methods, but automotive production demands extremely high accuracy – up to 30 cm in complex industrial environments with metal-rich settings, line-of-sight issues, reflections, and multipath effects. Achieving this precision is very challenging.

At the same time, UWB systems are costly and require extensive infrastructure for setup and maintenance.

Wi-Fi, with improved time-of-flight measurements, offers better accuracy than signal-strength-based methods, but in industrial settings its applicability is limited by interference and reflections. 5G thus became the next logical step, as it is increasingly being integrated into industrial applications.

Since environmental conditions can vary greatly depending on the use case, I first conducted a comprehensive use case analysis. These were then consolidated based on their requirements and transferred into three reference use cases, each defining different demands on the localization technology. Specifically, these are:

• “Find my production asset”: In this reference use case, various production assets such as pallets, forklifts, or large tools are localized in a relatively open environment free of machines.
• “Material flow”: In this reference use case, the flow of parts and production materials is tracked between the warehouse and assembly.
• “Process automation”: In this reference use case, vehicles and tools are localized in assembly to automatically control production processes. This environment is the most demanding of the tested areas.

Subsequently, areas within a plant were identified that correspond to these reference use cases.

My dissertation has shown that – depending on the use case and system configuration – sub-meter accuracies can in fact be achieved with 5G in the context of automotive production. This is a significant advance, as it underscores the technological feasibility of 5G-based localization systems in industrial environments.

Particularly remarkable was that the theoretical findings I had assumed earlier could be largely confirmed in practice. By classifying the requirements, I was able to demonstrate that these – depending on the specific use case – can actually be met.

However, it has also become clear that the introduction of 5G in industry is taking longer than originally expected. This is mainly because the 5G localization component represents an additional development step that requires further technological progress and adjustments.

This delay primarily affects the integration into existing infrastructures and the adaptation to industrial requirements. Many hardware components and industrially usable “off-the-shelf” solutions for 5G localization are still under development. Both manufacturers and standardization bodies are working to overcome these challenges and to further optimize them with a view to 6G.