Attention grounding: a new approach to in-vehicle information system implementation

Abstract

Driven by an expected $15-100 billion market, the race to produce in-vehicle information system (IVIS) functions has begun. Although IVIS functions may increase productivity, satisfaction and safety, they may also undermine safety and annoy drivers. Developing IVIS design strategies that minimize driver distraction is a critical challenge in developing successful IVIS functions. Several approaches have been developed to address this challenge. Interference mitigation has been the historical approach to IVIS research. More recently, workload management has emerged as an approach that may mitigate distraction by monitoring and managing the varying demands of driving and IVIS interaction. This paper presents attention grounding as a novel approach that complements previous efforts. Attention grounding uses the concepts of collaborative communication, grounding and dynamical systems theory to address the shortcomings of current approaches. IVIS interaction is considered a collaborative process that is supported with back-channel cues, rather than a series of discrete commands with no consideration of inevitable errors. Back-channel communication augments these commands to develop a shared awareness of the driver, roadway and IVIS state. Attention grounding considers the driver as an active participant in choosing when and how to use the IVIS, rather than assuming the driver's workload must be managed. This conceptualization highlights the role of dynamic changes in attention, such as attentional withdrawal and cognitive tunnelling as causes of distraction, rather than considering only mental overload. Together, these considerations provide a complementary approach to how IVIS might be designed to enhance ease of use and safety.

Role-Based Human-Computer Interactions

“…Sometimes, it may lead to loss of life, e.g., the mental overload of a pilot may compromise his/her command of the aircraft, and as a result, the aircraft may crash, resulting in casualties. That is why workload management is emphasized in in-vehicle information systems (IVIS) (Wiese & Lee, 2007). Currently, most research concerns static workload assessment, which comprises the assessment of workload prior to system application, and the design of the system by assuring that there is no overload.…”

Characterizing Driver Trust in Vehicle Control Algorithm Parameters

“…Using control-related cues to calibrate driver expectations, trust, and behavior is not a new concept (Walker, Stanton, & Young, 2006;Wiese & Lee, 2007); however, it has recently gained interest in the automated driving domain (Morando, Victor & Dozza, 2016;Toyoda, Domeyer, & Lenneman, 2017). Control-related cues arise from the motion or vibration of the vehicle because of control inputs like steering or speed control.…”

The challenge of designing intelligent support behavior : emulation as a tool for developing cognitive systems

“…Campbell et al, 2007;Green et al, 1995;Franzén et al, 1991;RESPONSE 3, 2009;Ross et al, 1995Ross et al, , 1996Stevens et al, 1999) such an approach does not make subsequent assessment redundant. Moreover, since the guideline based approach of design can be characterized as static (Wiese & Lee, 2007) evaluation becomes even more significant because guidelines cannot be referred to when several support behaviors might interact or when contextual demands might alter due to changes in driving context. While a commercially available and fully automated system for controlling a vehicle in traffic is not very likely for the near future -although testing autonomous vehicles is currently permitted in the US State of Nevada and a low-speed autonomous driving functionality is announced for the near future by several automakers (Newcomb, 2012) -the problems that might arise when the human driver needs to take over control from such a system were already addressed more than forty years ago by Sheridan (1970).…”