Involved Labs:

Cooperative Autonomous Systems (CAS), Institut für Angewandte Informatik und Formale Beschreibungsverfahren (AIFB), KIT
Modelling and Analysis in Mobility Software Engineering (MASE),  Kompetenzzentrum für Angewandte Sicherheitstechnologie (KASTEL), KIT
Socially Assistive Robotics with Artificial Intelligence (SARAI), Institute for Anthropomatics and Robotics (IAR), KIT

About the Project:

This interdisciplinary project investigates how humanoid social robots can serve as traffic mediators between automated vehicles (AVs) and vulnerable road users (VRUs). Within the project, SARAI leads WP3, focusing on the design of multi-modal interaction strategies that allow robots to communicate effectively with pedestrians in mixed-traffic environments. These strategies combine gestures, visual cues, and sound to deliver intuitive, understandable instructions—mimicking human traffic officers but enhanced with digital capabilities.
As part of this effort, SARAI conducted a user study to evaluate the effectiveness of canine-inspired motions on a quadruped robot in conveying emotional and intentional cues. Results show that motions intended to express alert, neutral, and yes/agree were recognized at above-average rates compared to prior studies on robotic emotional expression. Interestingly, prior experience with robots or dogs did not significantly impact recognition accuracy. These findings contribute to the development of expressive robot behaviors that are both intuitive and effective for real-world traffic interaction scenarios.
Current ongoing work builds on this foundation by extending the robot’s interaction capabilities through sound. We are implementing a real-time sound modulation framework on the robot that dynamically generates unique, non-repeating sounds synchronized with its motions. This extension aims to investigate whether combining acoustic and motion-based cues can further enhance the robot’s expressivity and communicative effectiveness in complex environments.