Communication Between Automated Vehicles and Drivers in Manual Driving Vehicles: Using a Mechanical Arm to Produce Gestures

Effective communication between automated vehicles and human drivers in manual driving vehicles is of great importance for traffic safety during the transition phase of automated vehicles. Gestures, which were widely used in road users' communication, were promising in conveying the intentions of automated vehicles naturally and intuitively without extra learning costs. However, the effect of gestures in conveying the automated vehicles' intentions on human understanding remains unknown. This study proposed the idea of adopting mechanical arms to produce gestures. An experiment based on video recordings was conducted to explore the effect of arms type (slow-waved mechanical arm (80 beats/min) vs. fast waved mechanical arm (120 beats/min) vs. human arm) and gesture type (taking the road vs. giving the road) on the participants' objective responses and subjective opinions. A total of 30 participants were recruited as human drivers in a manual driving vehicle, who received and responded to the gestures transferred by an encountering automated vehicle. Results indicated that regardless of the gesture type, the slow-waved mechanical arm led to a longer response time (mean +/- SD: 4.871 +/- 0.947 s) and lower response accuracy (88.3 +/- 32.4%) when compared with the human arm (response time: 4.457 +/- 0.727 s, response accuracy: 95.0 +/- 22.0%). It was also rated less understandable and comfortable than the human arm. Nevertheless, the fast-waved mechanical arm not only exerted as fast (4.484 +/- 0.818 s) and accurate responses (98.3 +/- 12.9%) as the human arm but was also rated as understandable, polite, and comfortable as the human arm. This indicated the implication of conveying gestures by utilizing the fast-waved mechanical arm (120 beats/min) to facilitate effective communication from automated vehicles to human drivers in manual driving vehicles. The present study's findings provided reference implications for manufacturers and designers to adopt this gesture-based communication method to develop safe and user-friendly automated vehicles.