Conference

9:15 a.m.
: Dr. Siegmar Haasis, founder and owner of Haasis DEC

The keynote presentation demonstrates how standardized data models, digital twins, and automated engineering processes are fundamentally transforming the development and industrialization of automotive wiring harnesses. Artificial intelligence opens up new possibilities—from automated wiring harness development and variant management to optimized production planning. At the same time, the momentum of Asian manufacturers is setting new standards for speed and efficiency: “China Speed” demonstrates how radically shortened development cycles are made possible through digitalization, platform strategies, and consistent automation. The keynote provides concrete insights into how companies can reduce complexity, significantly shorten development times, and secure their competitiveness in the age of software-defined vehicles.

9:45 a.m.
: Frank Syring, M.Eng., VMDS Product Owner at Volkswagen AG / Jorgos Kyriazis, M.Eng., VOBES Product Manager at Volkswagen AG.

VMDS (VOBES Master Data Service) is the new component management system for in-vehicle network development within the Volkswagen Group. The internal model is closely aligned with the VEC reference model, providing a mature modeling framework for component description and ensuring fundamental compatibility with numerous applications in the field of in-vehicle network development. A key requirement during the system’s design was the need to directly replace the legacy system: On the one hand, the new VOBES2025 toolchain had to be optimally supplied with data; on the other hand, backward compatibility with the existing VOBESplus toolchain and its individual applications had to be ensured. Another key focus was on the external supply of component data while avoiding duplicate data entry. In addition to an offline solution for data collection by second-tier partners, a VEC import interface opens up new opportunities for collaboration with component suppliers.

10:15 a.m.
Johannes Becker, Coordinator of the VES-WF and ECAD-IF project groups at PROSTEP IVIP e.V. and Managing Consultant at 4Soft GmbH

Model-based development of complex technical systems is increasingly taking place in a distributed and cross-organizational manner. In automotive electronics development in particular, OEMs, development partners, suppliers, and manufacturing facilities work on highly interconnected product models that are updated in short cycles. Today’s data exchange is largely based on the transfer of complete model states—so-called snapshots. This approach is inefficient, error-prone, and forces the recipient to manually identify the changes. Value already created is lost, responsibilities remain unclear, and true digital end-to-end traceability is prevented.

This presentation introduces the findings of the Vehicle Electrical Systems Workflow Forum (VES-WF) of the prostep ivip Association, which has developed an interoperable approach for the incremental exchange of design changes. The central idea: Instead of complete models, only semantically precise, machine-readable difference information between model versions is transferred—similar to patch files in software development. The presentation outlines the core requirements for such a format: unambiguous identification of model elements across system boundaries, precise and complete description of change content, traceability, and the detection of conflicts in parallel change processes.

One focus is on the evaluation of technological solution options. A comparison between XML- and RDF-based approaches shows that RDF offers significant advantages—particularly with regard to the mappability of model subsets, the ability to extend the model with change metadata through reification, and compatibility with existing ontologies. IRIs (Internationalized Resource Identifiers) are proposed as the identification mechanism, as they ensure global uniqueness, stability, and technology neutrality. A proof of concept validated the approach by converting VEC models to RDF, identifying graph-based differences, and applying them as changesets; the automated detection of change conflicts was also demonstrated in the process.

A key finding is that incremental change exchange is not a domain-specific issue, but rather a key requirement for all model-based development processes involving distributed collaboration. The presentation concludes with an outlook on the necessary cross-domain standardization and integration into existing toolchains.

11:15 a.m.
: Armin Hager, Founder and CEO of Aivato Inc.

The starting point is the observation that, while the now widely used best-cost engineering model for automotive electrical systems appears sensible from a cost-driven perspective, it is often inefficient in practice. Whereas in the past a single person was responsible for large parts of automotive electrical system development, the work today often consists of coordination, project management, and alignment among multiple best-cost locations—with the resulting communication and efficiency losses. At the AWH Congress in Detroit in October 2025, the AI technology vait (voitas AI technology) was unveiled for the first time; this technology uses artificial intelligence to fully automate specific design tasks in EDS engineering. The very positive response and the subsequent strong customer interest demonstrated that the topic is generating significant interest within the industry.

The presentation places this development in a broader context: It is assumed that the PC-based workplace will undergo fundamental changes over the next five to ten years—comparable to historic upheavals such as the mechanization of agriculture. The goal is to view this transformation not as a threat, but as an opportunity for greater productivity and better utilization of human capabilities. Finally, aivato will present an approach to accelerating this transformation via a central SaaS platform for in-vehicle network engineering. The focus is explicitly on technical added value, structural efficiency gains, and new development models—not on a product or sales presentation.

11:45 a.m.
, Nicola Wolter, Head of IT Consulting & In-Vehicle Network Technology at S-IT Application Engineering & Consulting GmbH.

The increasing complexity of electrical and electronic vehicle systems requires a consistent digital description from the system concept through to industrial implementation. In practice, however, there are gaps between conceptual and engineering data: system models are used only to a limited extent in downstream disciplines, while information from in-vehicle network development is incorporated into the system description only to a limited extent.

This paper addresses this gap and presents a concrete approach focused on the transition from system concept to vehicle electrical architecture. The goal is to create a consistent database for this phase, in which conceptual and engineering data can be developed in a traceable manner. The central question is how a structured description of electrical system connectivity can be derived from a SysML v2 system model and used as a reference for the vehicle electrical architecture. In particular, the analysis examines control units, sensors, actuators, their interfaces, and the logical connections between these system elements to ensure consistency between the system description and the vehicle electrical architecture.

The approach is not viewed as a purely one-way transformation, but rather as the foundation for closer integration between the system model and the vehicle electrical system architecture, in which changes are incorporated in a traceable manner and inconsistencies are identified early on. The physical implementation of the wiring harness—such as wires, connectors, or manufacturing parameters—is not part of this approach but is taken into account as a subsequent development stage. Using a prototypical approach, this paper demonstrates how system connectivity can be derived from SysML models and described in a structured manner in accordance with VEC to enable interoperable use within the vehicle electrical system architecture.

1:45 p.m.
: Dr.-Ing. Martin Weickgenannt, Managing Director of DiIT GmbH and Head of the Komax Group’s Competence Center for Smart Factory Solutions.

Wiring harness production is under increasing pressure due to a wide variety of variants, growing automation, and ambitious productivity targets. At the same time, the volume of available production and quality data is growing exponentially. The key question for decision-makers is therefore no longer whether to use data, but how to integrate it effectively and scalably into the production process. Drawing on the DAPROBO and Q1250 projects, this article demonstrates how Komax is implementing data- and AI-based approaches throughout the entire value chain of wire harness manufacturing.

DAPROBO (Data-Driven Productivity Boost) is a joint research and industry project between Komax and the Lucerne University of Applied Sciences and Arts (HSLU), funded by Innosuisse. The goal is to systematically analyze real-world machine data and use machine learning models to identify specific opportunities for optimization in terms of throughput, stability, and scrap. Based on historical and current production data, correlations between product variants, process parameters, and production results are analyzed and translated into clear, practical recommendations. Pilot applications demonstrate that significant productivity gains are possible without additional hardware investments, while at the same time making valuable expert knowledge available on a scalable basis.

With the Q1250 camera, Komax bridges the gap between image-data-driven analysis and directly adaptive machinery. The Q1250 exemplifies a new generation of intelligent monitoring systems in which AI-based models not only provide support but also actively contribute to quality assurance. Combined with the analysis and optimization approaches developed at DAPROBO, this creates a comprehensive concept: from the data-driven identification of optimal parameters to their consistent implementation in production and quality assurance. This article provides a strategic overview of DAPROBO and the Q1250 and demonstrates how data-driven optimization can be gradually integrated into existing production environments. For decision-makers in the automotive and Tier 1 sectors, it becomes clear how AI-based solutions deliver measurable added value, reduce risks, and enable a pragmatic transition to the next-generation smart factory.

2:15 p.m.
: Lukas Knak, CTO at Eisbach Robotik / Jan Olsson, CEO at Eisbach Robotik / Laura Roduner, CPO at Eisbach Robotik.

Automated wiring harness production is currently hampered by a significant gap between engineering and execution. While standardized design formats such as KBL and VEC provide detailed geometric and bill-of-materials information, an explicit, machine-readable production plan is lacking. Process planning therefore remains largely a manual, expert-dependent, and variant-specific effort that prevents economically viable high-variant production. Existing local optimizations and isolated automation steps merely address superficial symptoms without bridging the structural gap between engineering intent and production execution.

To address this challenge, we present a unified, software-driven architecture based on the Product-Process-Resource (PPR) paradigm and the Capability-Skill-Service (CSS) framework. The solution comprises an automated three-stage data refinement pipeline: first, a multimodal engine that uses Vision-Language Models (VLM) to convert text and visual layout data from fragmented sources into a Canonical Description Model (CDM); second, an algorithm that derives the physical resource layout and the Bill of Process (BoP) using expert-driven heuristics; and third, the matching of parameterized skills with the capabilities of robotic workcells, as well as data transfer via the OPC UA specification.

By using the Asset Administration Shell (AAS) as standardized information middleware, this approach enables the transformation of heterogeneous customer data into robot-ready manufacturing commands. The integration of standardized IDTA submodels creates a scalable foundation for cross-vendor production integration. This approach significantly reduces engineering effort and ramp-up times while simultaneously improving automation robustness and production efficiency. It thus provides a generalized methodology for the transition from manual planning to autonomous, interoperable control set manufacturing.

2:45 p.m.
: Dr. Alwin Hoffmann, Head of Industrial Digital Twin (AAS/DPP) Solutions at XITASO GmbH / Dipl.-Ing. Markus Rentschler, Research Coordinator for Digital Interoperability at ARENA2036 e.V.

This presentation provides an overview of the current state of standardized methods for creating the digital product passport and calculating the Product Carbon Footprint (PCF) for wiring harnesses in the automotive industry. The focus is on standardized data models for materials, processes, and emissions to enable transparent, automatable, and cross-company exchangeable CO₂ balances. The goal is to enable the industry to comply with upcoming legal requirements for creating the digital product passport with automated PCF calculation. The presentation will examine the existing AAS and Catena-X data models and highlight the challenges facing the manufacturing industry with regard to the necessary data and IT infrastructure.

3:45 p.m.
Daniel Dengel, Head of Business Development at fleXstructures GmbH.

Today, in-vehicle electronics development faces dual pressures for change: On the one hand, the variety of variants, the degree of integration, and time pressures in vehicle projects are continuously increasing; on the other hand, traditional, heavily manual development processes are increasingly reaching their limits in terms of robustness, scalability, and reproducibility. This presentation will demonstrate why physics-based design and validation of wiring harnesses and individual conductors must form the methodological foundation for a robust and fully digitized automotive wiring harness development process in the future.

What matters most here is not only the quality of the simulation, but above all the quality and standardization of the underlying data. Physically validated, certified cable and material data, provided centrally via a cloud-based platform, create—for the first time—a reliable “single source of truth” for development, simulation, and validation within the OEM environment. This makes it possible to break down internal silos, reduce the need for project-specific custom solutions, and ensure that results are reproducible across projects, platforms, and vehicle variants. As a result, standardized data structures and interfaces—particularly in the context of KBL and VEC—become key to automation.

The presentation highlights how KBL/VEC-based process chains, in conjunction with physically accurate models, directly support the automated design and validation of wiring harnesses and unlock significant efficiency potential. Using a real-world example from the automotive industry, the results are demonstrated: shorter development cycles, greater process robustness, and significant optimization potential. Specifically, the presentation examines automation use cases such as the complex optimization of multiple cables and hoses under changing boundary conditions, the automated optimization of grommets, and the static optimization of wiring harnesses based on predecessor derivatives—taking into account tolerances, clip connections, clearances to geometries, wire lengths, as well as misuse and incorrect installation scenarios. This approach yields cable length reductions of up to 10% in significantly less time. At the same time, it becomes clear that precisely this level of data quality is a prerequisite for future AI-supported design, optimization, and variant evaluation, since reliable AI results are only possible on the basis of physically accurate training and input data.

4:15 p.m.
: Dr.-Ing. Harald Bucher, Senior Product Manager for Electronic Logic and Wiring Harness Design at Vector Informatik GmbH.

The automotive industry is undergoing rapid transformation toward highly complex, software-defined vehicles, which requires new methods for efficient and reliable wiring harness development. Today’s environment is characterized by data scattered across various tools, fragmented responsibilities, and a multitude of inconsistent data formats—all of which disrupt data continuity and hinder collaboration along the value chain.

This presentation demonstrates how achieving data consistency across disciplinary and organizational boundaries—from architecture to routing design, between ECAD and MCAD, and at the OEM-Tier-1 interface—prevents information loss and enables consistent, model-based engineering. Consolidating distributed data into a single model-based approach supports horizontal and vertical traceability, automated routing synthesis, and the transfer of design variants from system architecture all the way through to routing and cable harness development.

OEMs face the challenge of managing variants within a unified platform, which requires a shift in perspective from predefined hardware variants to function-centric, generic variant descriptions—in line with the industry-wide transition from pre-manufactured hardware configurations to post-production software activation. Tier-1 suppliers must evaluate and validate Wire Harness under constant pressure to change, while managing thousands of updates during production and conducting concept evaluations, architectural studies, topology optimizations, and the counting and calculation activities required for RFQ processes. Together, these developments illustrate how data consistency serves as a critical enabler for managing complexity, improving collaboration, and accelerating development across the entire Wire Harness.