A Tangible Multi-Display Toolkit to Support the Collaborative Design Exploration of AV-Pedestrian Interfaces

Overview

• This paper presents a tangible multi-display toolkit to support the collaborative design exploration of autonomous vehicle (AV)-pedestrian interfaces.
• The toolkit allows designers and stakeholders to physically interact with and experiment with different interface designs in a shared environment.
• The goal is to better understand how these interfaces can facilitate communication and trust between AVs and pedestrians.

Plain English Explanation

The paper describes a new tool that can help designers and others create and test interfaces for autonomous vehicles (self-driving cars) to communicate with pedestrians (people walking). This is an important problem, as we need ways for AVs and pedestrians to understand each other and feel safe as self-driving cars become more common on our roads.

The tool they developed is a "tangible multi-display toolkit" - this means it uses physical, hands-on components that people can move around and interact with, along with digital displays, to simulate and test different interface designs. This allows designers to collaboratively experiment with various ways the AV could signal its intentions to nearby pedestrians, rather than just thinking about it on a computer screen.

By using this toolkit, the researchers hope to gain better insights into how these AV-pedestrian interfaces can build trust and facilitate clear communication between the two. The physical, collaborative nature of the tool is key, as it allows for richer exploration and faster iteration compared to traditional digital prototyping approaches.

Technical Explanation

The researchers created a tangible multi-display toolkit to support the collaborative design of AV-pedestrian interfaces. The toolkit consists of:

• A physical scale model of a road intersection with movable AV and pedestrian models
• Embedded digital displays to simulate AV interface designs
• Input devices like sliders and buttons to control the AV's behavior

This allows designers and stakeholders to physically interact with and experiment with different interface concepts in a shared environment that mimics real-world conditions. The goal is to better understand how these interfaces can facilitate communication and trust between AVs and pedestrians.

The toolkit supports both static displays (e.g. text, icons) and dynamic animations on the AV model to signal its intentions. Designers can also use the two-handed gestures and immersive sketching capabilities to quickly ideate and iterate on new interface concepts.

The researchers conducted a user study to evaluate the toolkit, finding that it enabled rich co-design experiences where participants could collaboratively explore and discuss AV-pedestrian interaction designs.

Critical Analysis

The paper provides a promising approach to addressing the critical challenge of designing effective communication between AVs and pedestrians. The tangible, collaborative nature of the toolkit allows for richer exploration compared to traditional digital prototyping methods.

However, the study was relatively small-scale, involving only 12 participants. Further research is needed to validate the toolkit's effectiveness with larger and more diverse user groups. Additionally, the paper does not address how the toolkit could be adapted for remote/distributed collaboration, which may become increasingly important as AV development involves geographically dispersed stakeholders.

Another potential limitation is that the toolkit only simulates a single intersection scenario. Expanding the toolkit to handle a wider range of road configurations and traffic situations could enhance its utility and real-world applicability.

Overall, the paper presents an innovative step towards improving AV-pedestrian interface design, but additional research and development will be needed to fully realize the potential of this approach.

Conclusion

This paper introduces a tangible multi-display toolkit that enables collaborative exploration of autonomous vehicle-pedestrian interfaces. By allowing designers and stakeholders to physically interact with and experiment with different interface concepts, the toolkit aims to provide richer insights into how these interfaces can facilitate communication and trust.

The results of the user study suggest that the toolkit can enable productive co-design experiences, but further research is needed to validate its effectiveness at scale and expand its capabilities to handle more diverse real-world scenarios. Nonetheless, this work represents an important contribution towards improving the design of AV-pedestrian interactions, which will be crucial as self-driving cars become more prevalent on our roads.