CREATING MILITARY SIMULATORS AT ARCADE GAME COST
George Katz
Scientific Management Associates SMA (Aust)
Sydney Australia
ABSTRACT
In these times of shrinking budgets and increasing equipment complexity,
training for major equipment acquisitions requires carefully selected,
cost effective life of type solutions. This paper briefly reviews
the types of solutions used to provide effective training for complex
real equipment. The introduction briefly outlines the advantages
and costs of different solutions. The body of the paper provides
a case study of a solution developed by SMA for a Royal Australian
Navy gun operator training requirement. Our approach has been to
exploit popular commercial off the shelf (COTS) hardware and software.
In particular, in the area of graphics, we chose to go in the direction
of Direct 3D (as opposed to Open GL) because of the benefits of
the parent Direct X suite of code, which supports sound, networking,
IO, as well as 2D and 3D acceleration. The paper concludes that,
although Part Task Trainers are not a simple panacea to all training
problems, advances in PC technology mean that they provide a highly
cost effective training solution - in fact at arcade game cost.
AUTHORS’ BIOGRAPHIES
Mr. George Katz, Lead Software Engineer, SMA Research
& Technology
Bachelor of Engineering (Computer Eng.) Honours
Since completing his degree in 1993, George has gained experience
in the use of several computer programming languages, including
Microsoft Visual C++, Borland Turbo C, Watcom C and various assembly
languages (68000; Hitachi SH-2; 68HC11). Prior to joining SMA George
worked in the USA for 2.5 years developing video games for Realtime
Associates. His experience included the design and implementation
of a multi-processor 3D game engine for Sega Saturn and for Sony
Playstation (Lead Programmer); design, development and implementation
of several computer games, including "Crusader: No Remorse" – Saturn
and Playstation (Co-lead Programmer); "NBA Live ’98" - Saturn (Lead
Programmer). George was also involved in the design and construction
of electronic and computer interactives for public exhibitions for
the Australian Museum from November 1994 – April 1995. He has also
constructed a number of mobile robots currently in use by PhD. students
at the University of NSW. George is currently leading the development
and integration of the software required to implement the two part
task trainers used as a case study in this paper. This task includes
working with and synthesising the efforts of the group’s educators,
hardware engineers, technicians and graphics designers.
Mr Andrew Wise, Manager, SMA Research & Technology
Bachelor of Engineering (Mechanical) Honours
Andrew has over nine years experience with the Royal Australian
Navy (RAN) as a Marine Engineer Officer. He therefore has a detailed
understanding of both the technical and operational requirements
of complex equipment within the Defence environment. Andrew joined
SMA as the Lead Instructor for Platform Systems for the RAN’s new
Minehunter project. He was subsequently promoted to Deputy Project
Manager for this program, and has most recently been promoted to
manage SMA’s Research and Technology Group. This Group designs and
develops technology based solutions for SMA’s international training
and Integrated Logistics Support projects. A particular focus of
the Group has been the development of a number of Part Task Trainers
and PC based simulations. These were developed to support the Minehunter
project, and two of these products form the basis of the case study
outlined in this paper.
CREATING MILITARY SIMULATORS AT ARCADE GAME COST
George Katz, Scientific Management
Associates SMA (Aust), Sydney Australia
INTRODUCTION
In these times of shrinking budgets and increasing equipment complexity,
training for major equipment acquisitions requires carefully selected,
cost effective life of type solutions.
One approach to this problem has been to embed sufficient training
functionality into the real equipment. This approach can add significant
cost to real equipment – a major budget driver where multiple units
are purchased. This also creates the need to trade off operational
time against training time.
A second approach is to use high fidelity simulation, such as full
motion simulators or training centres with a significant fit of
real equipment. However, this involves a high initial acquisition
cost and high consequential costs associated with software development/maintenance,
design-forced long lead times, and through life operation and maintenance.
Budget constraints often dictate that high fidelity simulators are
kept to a minimum. Where they do exist, they are often used for
non-core training purposes such as tactical development. They are
also frequently used for non-training purposes such as a system
test bed.
An alternative approach is to place more emphasis on the lower
fidelity/cost end of the training spectrum. This approach reduces
the loading on a high fidelity simulator or real equipment by ensuring
that trainees have previously mastered the basic skills. This can
allow use of the high-end simulator to focus on tactical or team
based skill development.
While traditional knowledge based learning supported by computer
based training, models, equipment parts and cutaways has its place,
it cannot prepare trainees fully for the tactile and spatial elements
of the task.
As a provider of training solutions for military projects, SMA's
challenge has been to develop training solutions that address these
issues. One approach has been the design of Part Task Trainers (PTTs)
to fill the gap in the training spectrum between traditional knowledge
based training and full simulators or real equipment.
Even within multi-billion dollar projects, the budget allocated
to training equipment is relatively small. Typically the budget
for even a particularly complex PTT is in the order of US$100,000.
Our approach has been to exploit popular commercial off the shelf
(COTS) hardware and software. We decided that the massive user base
and ongoing investment in continuing development by companies such
as Microsoft might work to our benefit. In particular, in the area
of graphics, we chose to go in the direction of Direct 3D (as opposed
to Open GL) because of the benefits of the parent Direct X suite
of code, which supports sound, networking, IO, as well as 2D and
3D acceleration. The benefits include the large established user
base, a degree of hardware independence, frequent updates, and backward
compatibility.
This approach has reduced coding by nearly 90% compared to ground
up development. It also enabled a successful progressive upgrade
of the hardware, Operating System and software with little or no
cost or schedule impact.
This paper presents a case study that expands on the nature of
the problem, the design solution applied and the success in achieving
the technical and commercial requirements.
BACKGROUND
In this paper we present two examples of Part Task Trainers (PTTs)
developed by SMA (Aust) as part of an articulated training solution
for a new class of minehunter being built for the Royal Australian
Navy. These PTTs emulate the 30mm Gun and the Electro-Optical Surveillance
System (EOSS). As a result of a detailed training analysis the PTTs
emerged as the most appropriate means to develop the identified
operator skills.
Importance of Human-Machine Interface (HMI)
The Human Machine Interface (HMI) of the real equipment is a key
factor in determining how effectively the overall system is used.
The operator manipulates the HMI to achieve the desired system outcomes,
and receives visual, aural and tactile feedback.
Therefore, regardless of the quality or utility of the real equipment’s
HMI, the PTT must replicate its features as accurately as possible.
This ensures that the trainee develops the physical and cognitive
skills necessary to use the system in its operational environment.
The fidelity of the PTT impacts directly the degree of learning
transfer that is achieved in the operational environment. The higher
the fidelity of the simulated HMI, the greater is the degree of
learning transfer that can be achieved.
Motivation Behind Development
Why would you choose to build a PTT? The main reasons are cost
and student throughput. A PTT cannot always replace a high end simulator
nor can it always eliminate real equipment training. Rather, it
often can be used to complement these.
The PTT is used to familiarise the student with the procedural
aspects of operating the piece of equipment. By the time the student
uses the high end simulator or the real equipment they are not wasting
time learning where a switch is or what it is supposed to do. Due
to the lower cost and task specificity of PTTs, training throughput
can also be increased.
Portability is also a major design criterion. Often it is difficult
to move a full motion simulator from site to site as training requirements
change. Both designs presented here took into consideration that
they must fit through a single door, run off a single standard power
point, and be transportable on an aircraft palette.
We will now give a brief description of the two PTTs and describe
what was done to keep development costs low, yet still provide an
effective training solution.
30mm GUN PART TASK TRAINER
Background
The entire training solution for the 30mm gun in the context of
the minehunter consists of a complementary suite of three components:
-
The 30mm Gun PTT, which will be used to train operators in
the aiming of the gun.
-
A real gun assembly and loader, which will be used for ordnance
training.
-
A gun simulation, which will be used for aspects of maintenance
training.
Our training specialists identified the required training outcomes
and we worked to achieve those (rather than our engineers simply
building what was possible). This meant that a significant amount
of time was spent on making the user interface look and feel virtually
identical to the real piece of equipment, rather than implementing
features that would have been desirable, but not necessary.
The gun has a relatively complex HMI (the reason the PTT was required).
Because the gun is stabilised only in elevation and the operator
is not, he has to constantly move his head to follow the gun sight,
which is attached to the stabilised gun. The operator has two handgrips
with switches and joystick, which he has to operate, as well as
a control panel and two dials to look at. He also wears a set of
headphones and a microphone, to receive fire control instructions.
The spatial relationships between the student and the significant
physical elements of the PTT were a key design criterion. This included
the student’s visual field, the physical location of switches, dials,
hand grips, seat position, etc.
|
