A practically developed approach to evaluate sonic interfaces of autonomous cars

Abstract: This thesis was a part of the SIIC-project (Sonic Interaction in Intelligent Cars) initiated by FFI (Fordonsstrategisk forskning och innovation), consulted by Volvo, Pole Position Production and Research Institutes of Sweden. The SIIC-project was about exploring sonic tools to affect the experience of self-driving cars. Sonic tools is implemented practically as sonic interfaces. Because of its visionary character, it was key to establish a foundation regarding future relevant information, user problems and user scenarios. The ideation process generated three conceptual sonic interfaces that accommodated three found user problems; Motion sickness, Informational overload and Low trust towards self-driving cars. One of these concept was made a functioning prototype implemented in video of relevant user scenarios. Carefully chosen metrics (Self- Assessment Manikin-, Van der Laan’s Acceptance- and Likert-scales) were assembled into an evaluation method. The method, together with the vide prototype, was then implemented in study with 30 participant, providing analytical material to finally review the metrics of the eval- uation method. The Self-Assessment Manikin and Van der Laan’s Acceptance scale were thought to highly contribute to the evaluation of participants experience of the interface. The Likert scales were less contributing, either deemed to be poorly implemented or simply not suiting for the cause. Main indications shows sonic interfaces seems to be capable of increasing a sense of trust towards the self-driving car and this capability seems to be evaluable.