Humanising Autonomy: Where Are We Going? is a book that calls for the people working towards autonomy – transport providers, OEMS, manufacturers, engineers, designers – to refocus their approach. We need to put human factors and users needs at the center of the conversation, rather than focusing on technological feasibility alone.
Some highlights below, of many!
Conversations & Perspectives: We spoke to people from different backgrounds, with different needs and unique opinions about their thoughts and hopes for the forthcoming ‘revolution’ of driverless transport.
Human–AV Interaction in the section "HOW WILL PEOPLE AND ROBOT CARS TALK?", explores the intricate ways self-driving vehicles (AVs) will interact with occupants and the external environment. Designing these interactions is significantly more complex than for generic robots, requiring a holistic approach.
The section categorizes AV interactions into five "faces":
• Inwards: Focuses on interactions with passengers, replacing the human taxi driver through modalities like speech or gesture. The authors note that in-car digital interfaces are already complex, combining audio, visual, tangible, haptic, and spatial tasks. Voice is highlighted for its broad communication bandwidth and accessibility.
• Outwards: Pertains to AV interactions with external parties such as pedestrians, other vehicles, and cyclists. "Social robotics" aims for robots that integrate into society and display social qualities, acknowledging human tendencies to anthropomorphize. Design principles emphasize balanced anthropomorphic projection (Design Principle 05) and establishing a human-robot relationship (Design Principle 06) through subtle cues like slowing down ("implicit interactions").
• Middleman: Describes the AV acting as an interface, communicating its observations, decisions, and intentions to occupants. This is crucial for semi-autonomous vehicles (SAE Level 2 or 3) due to the complex "handover" of control to the human driver. Building trust in new technology (Design Principle 04) is vital, as consumer distrust regarding AV safety is significant. Strategies like Volkswagen's "3+1" principle promote transparency.
The Inclusive design section emphasizes that automated systems, especially autonomous vehicles (AVs), must be designed for full inclusivity from the outset, rather than relying on retrofitted accessibility features that often segregate individuals with disabilities.
• Each Other: Focuses on AVs interacting with other robots to share data and learn. This machine-to-machine communication enhances safety and enables collective learning through techniques like "deep learning" and "reinforcement learning".
• Beyond: Examines how AVs integrate into a larger cloud infrastructure, contributing to a "cybernetic ecology" for efficiency and creativity in smart cities. This involves Vehicle-to-Infrastructure (V2I) and Vehicle-to-Everything (V2X) communication, and developing a "trust layer" based on transparency, orientation, value, and privacy. The book also stresses the importance of clearly communicating AV capabilities (Design Principle 10) and potentially removing user liability (Design Principle 11) for wider adoption.
Overall, the section stresses that effective AV design must address human needs and behaviors, advocating for observation over asking users what they want (Design Principle 13).
The section introduces Design Principle 07: Degrade Gracefully for the Senses, advocating for redundant communication channels (e.g., audio and visual) to accommodate varying sensory abilities. The Wayfindr project, an audio-based navigation standard for vision-impaired people, exemplifies how such inclusive solutions can universally benefit all users. Furthermore, it stresses mitigating technological biases in AI by using diverse training data to ensure AVs recognize all accents, languages, and skin tones. Ultimately, AVs should empower greater mobility for everyone.
Index of key sections & artwork provided by designers across the world
