Multitasking: a Review

NOVATEUR PUBLICATIONS
INTERNATIONAL JOURNAL OF INNOVATIONS IN ENGINEERING RESEARCH AND TECHNOLOGY [IJIERT]
ISSN: 2394-3696
VOLUME 2, ISSUE 6 JUNE-2015
1 | P a g e
Multitasking: a Review
Prof. Vyankatesh S. Kulkarni
Department of Mechanical Engineering, Solapur University /BIT/Barshi/India
Abstract
Multitasking (often referred to as timesharing) has been extensively studied from a mental workload and human
performance perspective. However, a relatively small amount of research has been conducted in the manufacturing
domain (Wickens, 1992). As the level of system automation increases, the role of the human has shifted from that
of a manual controller to system supervisor (Sheridan and Johannsen, 1976). According to Sheridan (1994),
“human operators in AMS make their way among machines, inspecting parts, observing displays, and modifying
control settings or keying in commands, most of it through computer-mediated control panels adjacent to various
machines.” This role of human operators in AMS has been identified as supervisory control in this paper.
INTRODUCTION
Human Supervisory Control
Supervisory control refers to one or more human operators programming and receiving information from a
computer that interconnects through artificial effectors and sensors to the controlled process or task environment
(Sheridan, 1987). Ammons, Govindaraj, and Mitchell (1988) described the supervisory controller as “an operator
responsible for a group of complex machinery where the operations require intermittent attention and depend on
higher-level perceptual and cognitive functions.” Sheridan (1976) defined a general paradigm of supervisory
control consisting of five functions: 1) Plan, 2) Teach, 3) Monitor, 4) Intervene, and 5) Learn. For each of the main
supervisory functions the computer provides decision-aiding and implementation capabilities, as shown in Figure 1.
A description of these functions is presented in Figure 2
Job scheduling, inventory planning, and problem solving (disturbance control) have been among the supervisory
control responsibilities commonly assigned to human operators in AMS (Suri and Whitney, 1984; Ammons et al.,
1988). The capabilities of humans and computers in AMS planning/scheduling tasks are presented in Table 1
(Nakamura and Salvendy, 1994). Table 2 shows examples of different types of unexpected contingencies
(disturbances) in AMS (Kuivanen, 1996). Ammons et al. (1988) stated that two ways in which the unique skills of
the human decision maker are used in supervisory control are to fine-tune or refine standard operating procedures
for particular system states and to compensate for unplanned events and unexpected contingencies.

ISSN: 2394-3696
2 | P a g e
Figure 1 General paradigm of supervisory control

Plan
a) model the physical system to be controlled
b) decide on overall goal or goals, the objective function, tradeoffs among goals,
And criteria for handling uncertainties
c) formulate a strategy or general procedure
Teach
a) Select the control action to best achieve the desired goal
b) Select and execute the commands to computers to achieve the goal
Monitor
a) allocate attention appropriately among the various subsystems to measure
Salient state variables
b) estimate the current state of the system
c) detect and diagnose any abnormality
Intervene
a) make minor adjustments of system parameters when necessary, as the
Automatic control continues
b)take over manual control if there has been a failure of the
c) abort the process in case of a major failure
Learn
a) develop understanding of and trust in the system
b) gain experience so as to do better next time
Figure 2 Temporal nesting of the general paradigm of supervisory control functions
VOLUME 2, ISSUE
model the physical system to be controlled
decide on overall goal or goals, the objective function, tradeoffs among goals,
criteria for handling uncertainties
formulate a strategy or general procedure
the control action to best achieve the desired goal
and execute the commands to computers to achieve the goal
allocate attention appropriately among the various subsystems to measure
estimate the current state of the system
detect and diagnose any abnormality
make minor adjustments of system parameters when necessary, as the
take over manual control if there has been a failure of the automatic control
abort the process in case of a major failure
develop understanding of and trust in the system
gain experience so as to do better next time
Temporal nesting of the general paradigm of supervisory control functions
ISSN: 2394-3696
3 | P a g e
decide on overall goal or goals, the objective function, tradeoffs among goals,
Temporal nesting of the general paradigm of supervisory control functions

ISSN: 2394-3696
4 | P a g e
Table 1 Capabilities of human and computer in planning/scheduling tasks of AMS

ISSN: 2394-3696
5 | P a g e
Table 2 Different type of disturbances in AMS

ISSN: 2394-3696
6 | P a g e
Determinants of Multitasking Performance
Different mental models have been used to describe multitasking performance. Scheduling, switching,
confusion, cooperation, and processing resources are mechanisms often identified as determinants of multitasking
performance (Damos, 1991; Adams, Tenney, and Pew, 1991; Wickens, 1992). In particular, the concept of
processing resources is the basis for understanding the other mechanisms, and hence for multitasking performance.
According to Wickens (1991), the resources concept is founded on the underlying assumption that the human
operator has a limited capacity for processing resources that may be allocated to task performance; therefore,
multitasking can lead to one or more tasks with less resource than required, causing performance deterioration. This
deterioration in the performance of one task because of competition with another task for critical resources is
known as interference.
Two major processing resources theories of task performance are Single-Resource Theory and Multiple-
Resources Theory. The Single-Resource theory proposed by Kahneman (1973) postulates one undifferentiated
limited pool of resources available to all tasks and mental activities. According to this theory, multitasking
performance declines as the difficulty of one of the tasks increases, because it demands more resources from the
limited pool, thus leaving fewer resources for performing the other tasks. Sanders and McCormick (1993) indicated
that the Single-Resource Theory has difficulty explaining: 1) why tasks that require the same memory codes or
processing modalities interfere more than tasks not sharing the same memory codes or processing modalities, 2)
why with some combinations of tasks increasing the difficulty of one task has no effect on the performance of the
others, and 3) why some tasks can be time-shared perfectly. According to Sanders and McCormick (1993), these
three issues can be explained by the Multiple-Resources Theory proposed by Wickens (1984).
The Multiple-Resources Theory proposes that there are three dimensions along which resources can be
allocated. The first dimension is stages (encoding and central processing vs. responding), which explains why tasks
requiring response selection and allocation resources are not disrupted by tasks requiring central processing
resources.
The second dimension is input modality (auditory vs. visual), which explains why multitasking is better
when the tasks do not require resources from the same modality than when they do. The third dimension is
processing codes (spatial vs. verbal), which explains why multitasking is performed better when one task involves
moving or positioning objects in space and the other involves language or logical operations. In addition to the
three dimensions mentioned above, this theory suggests a response dimension (vocal vs. manual), which explains
why multitasking is performed better when the tasks responses are of opposite types. Although the Multiple-
Resources Theory was developed based on dual task multitasking, it can be used to explain more complex
multitasking.
Scheduling and switching are highly influential on performance for both dual-task and more complex
multitasking. The operator’s scheduling and switching ability depends on an understanding of the temporal
constraints, the objective, and the cost associated with each task (Wood, 1982; Moray, Dessouky, Kijowski, and
Adapathya,1990). Poor scheduling, inefficient switching between tasks, or insufficient time to do the multiple tasks
sequentially will force the person to engage in concurrent processing. Wickens (1991) indicated that when the
operator is engaged in concurrent processing, multitasking performance will be influenced by: 1) confusion
(elements of one task become confused with the processing of another task because of their similarity), 2)
cooperation between task processes (caused by high similarity of processing routines), and 3) competition for task
resources. When the amount of resources demanded by the multiple tasks exceeds the amount of the operator’s
mental resources available, he or she will experience mental workload, consequently decreasing multitasking

ISSN: 2394-3696
7 | P a g e
performance (McCoy, Derrick, and Wickens, 1983; Bi and Salvendy, 1994). Figure shows the relationship among
the performance-resource function for multitasking (Wickens, 1992). Sheridan (1994) stated that mental workload
is very important for supervisory control in AMS where the human operator is constantly called upon to do multiple
complex sensory and judgmental tasks. The central issue for vigilance research is to determine the effect of the
additional tasks to the vigilance performance (Craig, 1991).
Conclusion
The paper reviews the evolution of multitasking capabilities of machines from the early rise of the concept of
multitasking and its benefits in today’s era.
References
[1] Advanced Manufacturing Systems: A Review Prof. Vyankatesh S. Kulkarni [IJIERT] ISSN: 2394-3696 VOLUME 2,
ISSUE 2 FEB.-2015
[2] Damos, 1991; Adams, Tenney, and Pew, 1991; Wickens, 1992
[3] Sanders and McCormick (1993)
[4] Unexpected contingencies (disturbances) in AMS (Kuivanen, 1996). Ammons et al. (1988)
[5] Sheridan and Johannsen, 1976
[5] McCoy, Derrick, and Wickens, 1983; Bi and Salvendy, 1994

Multitasking: a Review

More Related Content

Similar to Multitasking: a Review (20)

More from ijiert bestjournal (20)

Recently uploaded (20)

Multitasking: a Review