On-the-Move and in Your Car - Human-Computer Interaction and Innovation in Handheld, Mobile and Wearable Technologies - page 71

responses are considered to relate to the mental
workload and distraction associated with second-
ary tasks (Young, Regan and Hammer, 2003).
The advantage of this method over occlusion is
that it offers an assessment of cognitive, as well
as visual demand (of relevance to the assessment
of speech interfaces, for instance). The primary
disadvantage is that the method still requires
some form of driving task. Moreover, in contrast
with occlusion, the method has not been fully
standardised, and the ability to make cross study
comparisons is severely limited by the specific
choice of driving task scenarios (affecting task
load and the conspicuity of the peripheral stimuli).
It has also been noted that it is very difficult to
discern between the level of cognitive demand
and the visual demand for a given user-interface
(Young, Regan and Hammer, 2003).
An interesting recent development addresses
some of these limitations. Engstrom, Aberg and
Johansson (2005) considered the potential for
the use of a haptic peripheral detection task,
where drivers respond to vibro-tactile stimula-
tion through the wrist whilst interacting with an
in-vehicle system. Clearly, such a variation of
peripheral detection is not affected by variations
in lighting conditions. Furthermore, the authors
argue on the basis of their validation work that
this method provides “a “pure” measure of cogni-
tive load not mixed up with the effect of simply
looking away” (p.233).

ing system on a driver's awareness of the driving
environment (perception, reaction), and, their
ability to safely control the vehicle (manoeuvr-
ing, lane keeping)—Mattes (2003). Considerable
research is ongoing with the lane change task in
an attempt to develop an international standard
(Transport Canada, 2006). Key research issues
concern participant choice, training requirements
and developing acceptable limits for performance.

15 Second Rule

Participants carry out tasks with an in-car com-
puting system whilst stationary within a vehicle
or mock up (i.e. with no driving task) and with
full vision. The mean time to undertake a task is
considered to be a basic measure of how demand-
ing visually it is likely to be when driving (Green,
1999). A “cut-off” of 15 seconds has been set by
the Society for Automotive Engineers (SAE) - if
the task on average takes longer than 15 seconds
to achieve when stationary, it should not be al-
lowed in a moving vehicle. The method is simple
to implement and has the key advantage that it
has been formalised in an SAE statement of best
practice (SAE, 2000).
Research by Green (1999) and other research
teams (e.g. Pettitt et al., 2006) has shown strong
correlations between static task times and the total
amount of time spent looking away from the road
at displays/controls, both in simulator and road
studies. However, the correlation between static
task times and the duration of single glances
towards an in-vehicle display is generally poor.
This is important because a user-interface may
promote a small number of very long glances
(e.g. as a result of dynamically changing visual
information) which can have a considerable nega-
tive effect on driving performance (Burnett and
Joyner, 1997). It is for this primary reason that
many authors advocate the use of the occlusion
method as a better low-cost method for investigat-

Lane Change Task

This method occurs in a basic PC simulated envi-
ronment in which drivers are requested to make
various lane change manoeuvres whilst engaging
with an in-vehicle system. The extent to which
the profile of manoeuvre made by a driver var-
ies from the optimum manoeuvre (the normative
model) is considered to be a measure of the quality
of their driving. Specifically, the method has the
ability to assess the impact of an in-car comput-