CLICK BOND CHANGE YOUR HOLE APPROACH I WILL EXCEED YOUR EXPECTATIONS #06 RIVETLESS NUTPLATE CB6014 Exceeds NASM25027 strength requirements No. More. Holes Why is bonding better? CLICKBOND.COM/BENEFITS Researchers in Korea have collected data on how passengers interact with driverless vehicles to improve trust, which has been a challenge for the widespread adoption of autonomous vehicles operating at SAE L4 and L5, writes Nick Flaherty. Providing timely, passenger-specific explanations for vehicle decisions can bridge this trust gap, say researchers at the Gwangju Institute of Science and Technology (GIST). Existing explainable artificial intelligence (XAI) approaches cater to developers, focusing on high-risk scenarios, but this is unsuitable for passengers. The GIST team have developed TimelyTale, a dataset designed to capture real-world driving scenarios and passenger explanation needs. “Our research shifts the focus of XAI in autonomous driving from developers to passengers,” said Prof SeungJun Kim. The TimelyTale dataset includes information on the external environment, such as sights and sounds, proprioceptive data about the passenger’s body positions and movements, and interoceptive data about its sensations such as temperature. This data was gathered from passengers using sensors in various naturalistic driving scenarios to predict what explanations they would need. The dataset is publicly available for engineers to use. The research incorporates the concept of interruptibility, which is the shift in passenger focus from non-driving tasks to drivingrelated information. The visual explanations are combined with timing mechanisms that are constantly provided to be applied in risky scenarios. The team developed a machine-learning model that predicts the best time to provide an explanation and they carried out citywide modelling to generate textual explanations based on various locations. This found that visualising the vehicle’s perception state improves perceived usability, trust, safety and situational awareness without adding a cognitive burden on passengers. Driverless vehicles Discovering what passengers need to know in real driving scenarios AI algorithm boosts trust in driverless cars (Image courtesy of GIST)
RkJQdWJsaXNoZXIy MjI2Mzk4