Introduction
Automated and autonomous vehicles are on the rise and promise safer and more efficient mobility. But how do they communicate with other road users, especially pedestrians and cyclists? As part of the SALSA research project (Smart, Adaptive and Learnable Systems for All), we are working on the development and optimization of external human-machine interfaces (eHMIs). These novel communication options for self-driving vehicles are designed to enable intuitive and safe interaction with other human road users. In addition, the necessity of connecting external communication to the interior of the automated vehicle is also being considered.
The Objective of the eHMI Topic Area in SALSA
Our central goal is to develop a clear, comprehensible and universally usable communication method for automated and autonomous vehicles. We are researching how eHMIs can help automated and autonomous vehicles to clearly communicate their driving intentions and relevant information in order to avoid misunderstandings and potentially dangerous situations. To this end, studiokurbos is designing various solutions. In addition, we are looking at the interaction of eHMIs with other vehicle components and aspects. For example, the Hochschule der Medien (HdM) is looking at how explicit communication through eHMIs can be aligned with implicit communication through vehicle movement. At the University of Stuttgart, the eHMI is being examined from the perspective of vehicle design to ensure an aesthetic appearance. In addition, INVENSITY is examining the safety requirements for the communication systems. In order to evaluate different eHMI approaches, the German Federal Highway Research Institute (BASt) is developing the first approaches for an evaluation method for eHMIs. Audi is specifically focusing on the connection between external communication and the interior, and thus with the occupants of the autonomous vehicle. By taking a broad approach to the topic of eHMI, we are making an important contribution to the general acceptance, safety and user experience of autonomous mobility for all road users.
Approach and Methodology
In order to develop the interaction and communication of automated and autonomous vehicles with other road users in SALSA, research findings available at the HdM on behaviors in current and future road traffic were analyzed as possible models. To collect the previous research findings, the partners involved conducted literature research on interaction in today’s road traffic, on existing eHMI concepts and on problem scenarios in road traffic. The results were summarized in one-pagers and serve as a basis for the study design and further work in the project.
To develop effective eHMI solutions, we conducted a comprehensive analysis of existing and future communication relationships and selected two central use cases: intersection management and parking. These were categorized according to the criteria of information, intention and request in order to create a sound basis for a new communication model.
The methodological basis is the development of user journey maps and personas by studiokurbos to identify specific challenges and needs. For example, a blind pedestrian was developed as a persona to illustrate the requirements for barrier-free communication. This process is complemented by central workshops.
In current discussions among the partners, it has become apparent that, in addition to pedestrians, the focus for SALSA should increasingly be placed on cyclists as a target group, since they move faster than pedestrians in traffic. This poses particular challenges for communication with autonomous vehicles.
In addition to considering relevant use cases and developing use journeys, surveys are also being conducted to better understand the needs of the various users (including the occupants of the autonomous vehicle). Audi will also draw on its internal network and existing data. The University of Stuttgart is conducting a study with the aim of estimating the importance of the individual design features of eHMI. User surveys form the basis for further conceptual work, and monthly coordination rounds between partners promote holistic collaboration from a technical, psychological and design perspective.
To promote public trust and clear communication of eHMI, INVENSITY is conducting a comprehensive analysis of the safety of eHMI. The communication mechanisms between automated vehicles and other road users are being examined. Existing safety standards such as ISO 21448 (Safety of the Intended Functionality, SOTIF) and ISO/TR 4804 (Road vehicles – Safety and cybersecurity for automated driving systems) are analyzed to derive specific safety requirements and design recommendations. At the same time, recommendations for adapting the German Road Traffic Regulations (StVO) are developed to create the legal framework for the integration of eHMI.
Research Approaches and Solution Development
Based on the results of the use case analysis, we are currently in the design phase to develop specific display and operating concepts.
studiokurbos has already developed initial conceptual proposals for the visual categories of form, color and light. These serve as a basis for further discussions and optimizations.
The literature analyses were supplemented by a qualitative observation and interview study conducted by the HdM in San Francisco. There, so-called robotaxis from the provider Waymo are already integrated into road traffic across the board. The user requirements and needs of road users in mixed traffic were examined.
Visualization and Prototyping
To illustrate the concepts, the studiokurbos designers create 3D visualizations using Blender. The visualizations serve not only as illustrations, but also as a basis for further discussions with the partners.
In addition to visualizations by studiokurbos and low-fidelity prototypes by the HdM, the concepts for combining exterior (light) communication and interior space are also to be implemented as prototypes by Audi. Based on this, studies for evaluation are planned, implemented and evaluated.
Evaluation
The eHMI low-fidelity prototypes are evaluated in an empirical user study in a driving simulator and further developed in a user-centered way. In subsequent studies under scientific standards, the prototypes are consistently improved for the experience of external road users by adapting the study conditions to those of a real situation.
With the designs of studiokurbos, several studies are planned at the University of Stuttgart to evaluate the concepts. For this purpose, a VR driving simulator is also used at the University of Stuttgart, which enables an immersive experience of the concepts in realistic traffic conditions. The findings from the user studies are used to iteratively revise and optimize the concepts.
Using a test subject study, BASt is testing initial approaches for an evaluation method for eHMIs in order to be able to evaluate eHMI concepts for communicating with other road users in the future.
Integration into the SALSA Project
Our research approaches in the field of external communication are closely linked to the goals of the SALSA project. By improving human-machine interaction, we contribute to increasing traffic safety, comprehensibility and acceptance of autonomous vehicles. The interaction with the HMI offers in the vehicle forms the bridge between inside and outside, between interior and exterior. The developed concepts and visualizations are continuously being further developed and discussed in collaboration with the project partners to enable optimal implementation in real traffic.
Conclusion and Outlook
Communication between autonomous vehicles and non-autonomous road users is a central element for the successful integration of autonomous mobility. Our ongoing research within the SALSA project lays the foundation for innovative eHMI concepts that strengthen transparency, trust, and safety in road traffic. The resulting concepts and study results are presented to the public at the SALSA mid-term and final events, as well as continuously through our channels.
We thank all partners for their support and look forward to the next steps in the SALSA project!
