Using CWA to Improve the Situation Awareness of Underwater Remotely Operated Vehicle Pilots

1
Using CWA to Improve the Situation Awareness of
Underwater Remotely Operated Vehicle Pilots

• Peter Henley
• Centre for Marine Science and Technology
• Curtin University

2
Presentation Outline

• Background to the research
• ROV and umbilical modeling
• Why CWA?
• ROV work domain analysis
• Constraint-based task analysis
• Display design

3
Personal Background

• Marine Engineer (Steam)
• Degree in Marine Physics at UCNW, Bangor, UK
• Commenced working with ROVs in Aberdeen in 1980
• 20 years experience in the supervision and
  project management of ROV operations in the
  offshore oil and gas industry

4
(No Transcript)
5
Problems Faced by ROV Pilots

• The Underwater Environment
• Relative magnitude of forces
• EM radiation absorption scattering
• Acoustic system problems
• Inadequate Information
• Ad-hoc sensor selection
• Commercial pressures
• Poor Quality User Interfaces

6
(No Transcript)
7
Research Goal

• To develop a single-screen situation awareness
  display for operational use by ROV pilots
• What information is needed?
• How can this information be derived ?
• How can this information best be presented?

8
Endsley Model of SA and Naturalistic Decision
Making

• Endsley (1988)
• the perception of the elements in the
  environment within a volume of time and space,
  the comprehension of their meaning, and the
  projection of their status in the near future.
• Three types of cognitive behaviour
• Automaticity
• Schema-based recognition-primed decision making
• Linear information processing

9
(No Transcript)
10
Initial Task Analysis Methodology

• Expert interviews with experienced ROV pilots
  (minimum 2 years experience)
• Focus-group type of interview and free-flowing
  discussion encouraged
• Interviews framed by particular ROV mission tasks
• Mission tasks broken down into subtask goals that
  were common across a range of missions
• Notes made at interview and all interviews taped
• Data sought was the SA information sources
  actually used by operators to carry out
  individual piloting tasks/actions

11
Interview Data Analysis Problems

• Interviewees regularly made the comment that
  every instance of a particular task is different
  because of environmental conditions, access
  constraints and other factors outside the control
  of the operators
• The interviewees continuously made suggestions
  for instrumentation and information sources that
  were not generally available to ROV pilots

12
Interview Data Analysis Problems

• Because of their high skill levels, almost all of
  the pilots interviewed were employed operating
  large, powerful work-class vehicles which cope
  more easily with adverse environmental conditions
  than the smaller, observation-type vehicles
• It became apparent that neither the task analysis
  attempted nor the interface design
  recommendations made by Endsley and others were
  able to specify exactly what information to
  include in the interface

13
Methodology Acceptance Criteria

• The methodology must be capable of specifying
  direct support for the three generic behaviour
  types identified from the Endsley SA and
  decision-making model.
• The methodology must be capable of defining
  information support requirements for all levels
  of operator ability and experience.
• The methodology should result in information
  support recommendations that are not at variance
  with the general guidelines for support of the
  three levels of SA derived from the Endsley
  model. 

14
Methodology Acceptance Criteria

• The methodology must be capable of a clean-slate
  approach that can identify a minimally complete
  SA information set.
• The methodology must be capable of identifying
  the sources of necessary SA information.
• The methodology must be able to examine known
  tasks in such a manner as to extract generic
  information support requirements that are
  applicable across the broadest possible range of
  common subtasks and task-types.
• The methodology must be able to examine all
  aspects of environmental influence on an ROV
  system.

15
(No Transcript)
16
Control Task Analysis

• Constraint-based analysis on interview data
• For each control-task, identified inputs, outputs
  and any constraints on pilot actions
• Final product was a list of the intrinsic work
  constraints common to all control tasks

17
Task Analysis Summary

• 11 ROV missions examined
• 14 sub-mission goals identified
• 26 control tasks analysed
• 20 operational constraints derived

18
Display Design

• Interface requirements defined in terms of
  constraint proximity monitoring and alerts
• Information framework and grouping from WDA
• SA information sensors and modeling requirements
  from WDA
• Rasmussen/Vicente SRK taxonomy
• Display type, geometry and viewpoints adopted
  from avionics literature

19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
Possible Future Work

• A Position Estimator is required to provide
  faster screen updates than will be permitted by
  the sensor sample rates
• A GIS Mapper is required to accept both existing
  and new bathymetric data and objects/structures
• A Sonar Target Extractor is required to extract
  and present targets to the display and the GIS
  Mapper
• A Dynamic Positioning System is required to
  allow the ROV to automatically hold a mid-water
  position on command.

26
Acknowledgements

• Defence Science and Technology Organisation -
  Maritime Platforms Division
• Dr. Roger Neill
• Curtin University Prof. John Penrose
• Prof. Geoff
  West

27
(No Transcript)