Jaemin Chun

Vibrotactile Feedback for Information Delivery in the Vehicle

As technology advances, more functions have been, and continue to be added to the vehicle, resulting in increased needs for improved user interfaces. In this paper, we investigate the feasibility of using vibrotactile feedback for in-vehicle information delivery. First, we measured the spectral characteristics of ambient vibrations in a vehicle, and designed clearly distinguishable sinusoidal vibrations. We further selected via dissimilarity rating the four sets of sinusoidal vibrations which had three to six vibrations. Second, we evaluated the learnability of the vibration sets when associated with common menu items of a Driver Information System (DIS). We also replaced the two most confused sinusoidal vibrations with patterned messages, and assessed the degree of learnability improvement. Finally, we evaluated the extent to which participants could select a desired function in a DIS via vibrotactile messages while simultaneously performing a driving-like primary task with higher priority. The results demonstrated high potential for vibrotactile messages to be effectively used for the communicative transfer of in-vehicle system information.

Sensors Know When to Interrupt You in the Car: Detecting Driver Interruptibility Through Monitoring of Peripheral Interactions

Interruptions while driving can be quite dangerous, whether these are self-interruptions or external interruptions. They increase driver workload and reduce performance on the primary driving task. Being able to identify when a driver is interruptible is critical for building systems that can mediate these interruptions. In this paper, we collect sensor and human-annotated data from 15 drivers, including vehicle motion, traffic states, physiological responses and driver motion. We demonstrate that this data can be used to build a machine learning classifier that can determine interruptibility every second with a 94% accuracy. We present both population and individual models and discuss the features that contribute to the high performance of this system. Such a classifier can be used to build systems that mediate when drivers use technology to self-interrupt and when drivers are interrupted by technology.

Exploring the Value of Information Delivered to Drivers

New IT functions have greatly increased the amount of in-car information delivered to drivers. Although valuable, that information can distract drivers when delivered during vehicle operation. By inferring driver state from sensor data, prior research has shown that it can accurately identify opportune moments to deliver information. Now that we know when to best deliver information, it raises the question: what information should we deliver at those interruptible moments? To answer this question, we conducted a series of surveys and interviews and compiled a list of representative in-car information items and context factors that affect the importance of these items. By combining and exploring those context factors, we identified driving situations when each of the in-car information items is highly valuable, and verified these situations through a large online survey of drivers. Lastly, we examined what technology is available for detecting these driving situations, and which situations require further advanced technologies for detection. Results from our study offer important insights for understanding the diversity of drivers’ experiences about the value of in-car information and the ability to determine situations in which this information is valuable to drivers. With these results, researchers can then build information delivery systems that can deliver information to drivers both when they are interruptible and when they find the information valuable.