A Collaborative Crowdsourcing Method for Designing External Interfaces for Autonomous Vehicles

TL;DR

The paper tackles the scalability gap in participatory design for external interfaces of autonomous vehicles by introducing a collaborative crowdsourcing method that pairs crowd creativity with expert feedback. Using iterative idea building and a tree visualization, everyday participants generate eHMI concepts that are refined across rounds, guided by expert assessments of effectiveness and feasibility. Final evaluation compares a popular crowdsourced design, an innovative crowdsourced design, and a baseline design through video based simulations, measuring interpretation speed and user experience. Findings show that familiar, standardized signals yield the best interpretability and UX, while innovative concepts remain competitive, illustrating the potential of scalable, participatory methods for shaping emerging technologies.

Abstract

Participatory design effectively engages stakeholders in technology development but is often constrained by small, resource-intensive activities. This study explores a scalable complementary method, enabling broad pattern identification in the design for interfaces in autonomous vehicles. We implemented a human-centered, iterative process that combined crowd creativity, structured participatory principles, and expert feedback. Across iterations, participant concepts evolved from simple cues to multimodal systems. Novel suggestions ranged from personalized features, like tracking lights, to inclusive elements like haptic feedback, progressively refining designs toward greater contextual awareness. To assess outcomes, we compared representative designs: a popular-design, reflecting the most frequently proposed ideas, and an innovative-design, merging participant innovations with expert input. Both were evaluated against a benchmark through video-based simulations. Results show that the popular-design outperformed the alternatives on both interpretability and user experience, with expert-validated innovations performing second best. These findings highlight the potential of scalable participatory methods for shaping emerging technologies.

Figures (11)

• Figure 1: Illustration of the Expert Analysis and Feedback process. Participant-submitted sketches are grouped into similar clusters (concepts). An independent expert rates each concept's effectiveness and feasibility, providing justifications. These illustrated concepts, with descriptions and feedback, form the next stage of the visualization tree, guiding the next iteration.
• Figure 2: Illustration of the Crowdsourced Design Collaboration process. Participants iteratively refine ideas inspired by earlier contributions. Using a tree-based visualization, top-level concepts branch into subsequent sketches. Participants review and rate prior concepts and then submit their contribution. Each submission includes a sketch and a brief explanation.
• Figure 3: Configuration of the traffic scenario used for evaluating eHMI designs. Key points where specific cues were activated are highlighted: (Point A) communication of pedestrian detection, (Point B) signaling yielding intention, and (Point C) additional crossing facilitation. Sample screenshots from the simulation of an autonomous bus without eHMI are shown at the bottom.
• Figure 4: Concepts developed in the first collaborative iteration. All concepts have a brief descriptions and expert feedback, along with expert ratings for effectiveness ($\circ$) and feasibility ($\diamondsuit$), where 7 indicates the highest effectiveness and feasibility.
• Figure 5: Concepts developed in the second, third and fourth iterations. All concepts have brief descriptions and expert feedback, along with expert ratings for effectiveness ($\circ$) and feasibility ($\diamondsuit$), where 7 indicates highest effectiveness and feasibility.