Academic journal article Human Factors

Speech-Based Interaction with In-Vehicle Computers: The Effect of Speech-Based E-Mail on Drivers' Attention to the Roadway

Academic journal article Human Factors

Speech-Based Interaction with In-Vehicle Computers: The Effect of Speech-Based E-Mail on Drivers' Attention to the Roadway

Article excerpt


The evolution of computers has made a variety of in-vehicle information systems possible (Lee, 1997; Lee, Kantowitz, Hulse, & Dingus, 1994). These new information systems can enhance mobility and productivity, but they may also distract drivers and undermine safety. Automatic speech-recognition (ASR) and text-to-speech technology enable a new driver-vehicle interface that could potentially enhance safety: A speech-based interface allows drivers to keep their hands on the wheel and eyes on the road. Some designers assume that interacting with an in-vehicle computer through ASR and text-to-speech technology will not distract the driver. This assumption ignores the potential cognitive load of speech-based interactions and its effect on driving performance. Little research has addressed the cognitive load of a speech-based interface for in-vehicle computers, and none has examined its effect on driving performance.

Attention to the Road and Speech-Based Interactions

Researchers and designers have long recognized the potential of visual displays to distract drivers (Lunenfeld, 1989; Mollenhauer, Hulse, Dingus, Jahns, & Carney, 1997; Srinivasan & Jovanis, 1997; Wierwille, 1993), but few have addressed the possibility of auditory displays and verbal controls producing similar effects. Although speech-based interfaces have not been examined directly, substantial research has considered the effect of telephone use on driving performance and safety. Comprehensive reviews suggest that speech-based interaction has the potential to distract drivers and degrade safety (Goodman, Tijerina, Bents, & Wierwille, 1999; McKnight & McKnight, 1993; Parkes, 1993). For example, analysis of crash data has shown that drivers using handheld cellular telephones experience at least a fourfold increase in crash risk and that hands-free telephones may not provide a large safety benefit (Redelmeier & Tibshirani, 1997; Violanti, 1997). A large number of on-road and simulator studies have shown subst antial increases in driver response times during telephone conversations (Alm & Nilsson, 1995; I. D. Brown, Tickner, & Simmonds, 1969). A recent study directly compared the visual demands of entering numbers into a keypad with the cognitive demands of a memory and addition task. The results showed that both tasks impaired drivers' ability to detect braking vehicles by 500 ms (Lamble, Kauranen, Laakso, & Summala, 1999). Speech-based interaction with in-vehicle computers is similar to cellular telephone conversations, and so a poorly designed interface may distract drivers.

The distraction and safety consequences of speech-based operating systems are not easily extrapolated from research on speech communication with a cellular telephone or other concurrent verbal tasks. Talking to a computer is fundamentally different from talking on a cellular telephone or to a passenger. On one hand, some differences include the added cognitive load of interpreting a poor-quality synthetic voice (Smither, 1993) as well as that of producing speech that can be recognized by the system (particularly in noisy environments). Recalling commands, remembering system syntax, and recovering from errors could also add a substantial demand not experienced in conversations. Similarly, the spatial demands of navigating a complex menu structure may distract some drivers in ways that a conversation would not (Vicente & Williges, 1988). On the other hand, interaction with a speech-based system may be less distracting than conversations, given that drivers may have more flexibility to abort an interaction with an in-vehicle computer -- there is no social imperative to continue a conversation with a computer. Concerns about the safety of cellular telephone conversations and the unique characteristics of a speech-based interface make an empirical investigation of speech-based interfaces critical.

The distraction potential of a speech-based interface may depend on its complexity. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.