In-vehicle human-machine interaction (HMI) requires varying degrees of visual and cognitive resources. Concerns over excessive visual demands in the vehicle have existed for some time. More recently concerns over the impact of HMI on drivers’ cognitive resources have gained attention. While multiple definitions of cognitive load (also called cognitive or mental workload) appear in the literature, it is commonly defined as the relationship between the cognitive demands of a task and the cognitive resources of the user. A central question in designing HMI for in-vehicle devices is how the HMI will impact the driver’s cognitive load. In-vehicle devices are often operated while the vehicle is moving. While the primary task of the driver is to ensure driving safety, the availability of such devices often lures drivers into getting engaged in peripheral tasks while driving. Poorly designed HMI requires an increased level of cognitive resources, reducing the driver’s ability to dedicate sufficient cognitive resources to the driving task, and can lead to possibly disastrous consequences. The Yerkes-Dodson Law provides a theoretical background for modeling the effect of driver cognitive load on driving performance, and can be seen as a pivotal concept in the detection and management of cognitive load. While research results on in-vehicle cognitive load are frequently presented at automotive research conferences and in related journals, CLW 2012, the second in the series, will provide a unique forum for focused discussions on this topic.
The workshop has four goals:
- Explore the concept of cognitive load: While the concept of cognitive load has been used by a number of researchers working on in-vehicle HMI (as well as those working in other fields), the definition of cognitive load sometimes seems illusive. What exactly is cognitive load? The workshop will explore different points of view on this question.
- Explore issues in cognitive load estimation: Estimating cognitive load while driving is a challenging task. Clearly, our understanding of estimation is tightly coupled to our definition of cognitive load. However, whatever the definition we use, estimation (on-road, and laboratory-based), focuses on three types of measures: performance, physiological and subjective. The workshop will explore the practical use of these measures in on-road studies and those performed in a laboratory setting (both using immersive driving simulators and other techniques).
- Explore issues in cognitive load management: How can we design in-vehicle HMI such that the driver has the cognitive resources to safely operate the vehicle, even while interacting with in-vehicle devices? Researchers and practitioners have explored a number of approaches for workload management, from simply turning off HMI in certain situations, to introducing novel interaction methods which hopefully do not introduce undue cognitive interference with the driving task (voice interfaces, augmented reality, mediation, tactile interfaces, subliminal notifications, etc.). Other work suggests that effective implementations of these and other systems need to adapt to the driver’s state. The workshop will explore various aspects of managing the driver’s cognitive load.
- Explore paths for future research and development: In light of current approaches to cognitive load estimation and management, what research and development avenues should be explored in the next 2-10 years? Workshop participants will discuss (a subset of) problems to be explored, goals to be set, hypotheses to be tested, and approaches likely to be fruitful in testing these hypotheses.
The workshop organizers will bring together a number of experts from government, industry, and/or academia to address topics on exploring the concept of cognitive load (goal 1).
Furthermore, we solicit research papers exploring issues in cognitive load estimation and management for interactions with in-vehicle devices (goals 2 and 3). Authors are encouraged to also include at least one paragraph addressing paths for future research and development (goal 4). Additionally, position papers on goal 4 are solicited. Topics of interest include:
- Cognitive load estimation in the laboratory,
- Cognitive load estimation on the road,
- Sensing technologies for cognitive load estimation,
- Algorithms for cognitive load estimation,
- Performance measures of cognitive load,
- Physiological measures of cognitive load,
- Visual measures of cognitive load,
- Subjective measures of cognitive load,
- Methods for benchmarking cognitive load,
- Cognitive load of driving,
- Cognitive overload and cognitive underload,
- Approaches to cognitive load management inspired by human-human interactions.