(i) SOCIETY FOR

PREPRINT
NUMBER
MINING, METALLURGY,
AND EXPLORATION, INC.
94-114
P.O. BOX 625002 LITTLETON, COLORADO 80162-5002
A PC BASED ERGONOMIC ANALYSIS SOFTWARE PACKAGE
FOR UNDERGROUND MINING EQUIPMENT
R.L. Unger
US Bureau of Mines
Pittsburgh, Pennsylvania
For presentation at the SME Annual Meeting
Albuquerque, New Mexico - February 14-17, 1994
Permission is hereby given to publish with appropriate acknowledgments, excerpts or
summaries not to exceed one-fourth of the entire text of the paper. Permission to print in more
extended form subsequent to publication by the Society for Mining, Metallurgy, and Exploration
(SME), Inc. must be obtained from the Executive Director of the Society.
If and when this paper is published by the SME, it may embody certain changes made by
agreement between the Technical Publications Committee anp the author so that the form in
which it appears is not necessarily that in which it may be published later.
Current year preprints are available for sale from the SME, Preprints, P.O. Box 625002,
Littleton, CO 80162-5002 (303-973-9550). Prior year preprints may be obtained from the
Engineering Societies Library, 345 East 47th Street, New York, NY 10017 (212-705-7611).
PREPRINT AVAILABILITY LIST IS PUBLISHED PERIODICALLY IN
MINING ENGINEERING
Abstract. Restricted fields of vision are a
common problem with underground mining
equipment. Also, when lighting systems are provided,
they may be positioned where they cause glare. This
restricts the visibility around the machines even
further. To help address these problems, the U.S.
Bureau of Mines has developed a PC based software
package to aid in the analysis of certain human
engineering aspects of mining equipment design. The
software is available for use by equipment
manufacturers and mining companies for design work
on new machines, as well as for evaluating proposed
modifications to existing machines.
Ergonomic Design Problems in Underground Mining
Designing operator compartments (crewstations)
for underground mining equipment can be a
formidable task. Massive machinery is required to
meet production goals while the confined environment
imposes severe space restrictions in every direction,
most critically in height. As a result, mining
equipment operator compartments are frequently
smaller than the recommendations to adequately
accommodate most of the population. In fact,
crewstations which provide operator accommodations
of less than 0.75 meters in height and 0.60 meters in
width are not uncommon .. In these cramped quarters,
visibility is often severely impaired, forcing operators
to lean outside of the protection of their crewstations
in order to see. When lighting systems are provided,
they may be positioned where they cause glare, which
further restricts the visibility of those working on or
around the machines. Critical controls are often
placed where they are difficult for some operators to
reach in the awkward seating postures which they
must assume. In panic situations, the wrong control
may be activated, which can lead to an accident.
To effectively meet the challenges of underground
equipment design, there are a variety of human factors
issues manufacturers should address. However,
research indicates that this is often not the case. A
study of the design practices of 44 engineers
conducted in the early 1970's concluded that: "One of
the most consistent findings of our research,
confirmed in each study with all subjects regardless of
sophistication, years of experience, or design problem,
is that the typical design engineer does not consider
human factors in his design" (Meister, 1971).
Although this study was conducted twenty years
ago, the problem still exists today. An informal
survey conducted in 1986 at a Bureau sponsored
seminar on mine equipment design found that none of
the participating design engineers from a major mining
equipment manufacturer had ever received formal
training in either the design process or in human
factors design. In fact, most admitted overwhelming
reliance on designs that had been used before rather
than on development of novel approaches.
Given the mining industry's lack of expertise in
human factors engineering, the Bureau has been
attempting to provide it with products to assist in the
1
equipment design process. Several recommendations
documents have been developed, including a
maintainability design reference and a textbook on
human factors in mining. In addition, a database
containing abstracts of human factors research
applicable to mining was developed and made available
to anyone with a Digital Equipment Corp. VAX
compatible terminal and modem. One of the most
recent Bureau research projects that shows promise of
impacting how engineers think about the mining
equipment design process is the development of the
Crewstation Analysis Programs (CAP) package. CAP is
a set of computer programs that can be used to analyze
some of the human engineering aspects of crewstation
design that have particular significance in underground
mining. These programs currently include assessments
of visibility, illumination, and glare in the surrounding
work area. Graphics oriented procedures allow the user
to input information needed for the analyses, such as
type of mining machinery, lighting systems and mine
layouts. Once this information is identified, the user
may select any of the available analyses. By
automating the task of evaluating human factors design
issues, CAP will allow 'engineers to more quickly and
easily experiment with novel crewstation layouts for
mining equipment.
The Design of CAP
Development of the CAP package has been ongoing
for several years. The original version of the software,
completed in 1990, was written in FORTRAN for a
Digital MicroV AX IT computer, and required a
Tektronix graphics display terminal and mouse, a
graphics tablet, and an optional hard copy unit. The
programs have recently been updated and translated to
the C language and are now targeted for the widely used
386/486 processors. However, the programs have been
designed so that they can be easily transported to other,
more powerful platforms, such as workstations. A great
deal of effort was expended to develop interactive
graphic displays that are "user friendly" to assist in the
placement of machines, lamps and other objects needed
to perform the various ergonomic analyses.
CAP is designed much like a simple Computer
Aided Drafting (CAD) program, such as AutoCAD.
The center of the screen is taken up by a large viewport
into the "world", where the user positions 3-dimensional
models of mines, machines, lamps, glare shields, or any
object that is needed by the analysis (Figure 1). The
lower right portion of the screen contains the viewing
controls, which allow the user to change their point of
view with respect to the objects in the scene. These
controls are always available and activate several
zooming, panning and camera positioning functions.
The lower left portion of the screen contains controls
for the activation of utilities and positioning aids. These
include changing the color or visibility of objects in the
scene, turning on a grid to help position objects,
toggling the scene from a wireframe to a solid mode,
and activating the help screen. '
The upper right portion of the screen contains the
2
14:11 VIEW: REAR RIGHT ZOOM: 1.250 Cur Mach: BATSCOOP SNAP HN HIDE
Ill.UM
VIEW
PRINr
CLRIll.
SHOW
SHADOW
ll..AMP
ELAMP
VPANEL
HPANEL
EPANEL
CLRALL
OPEN
SAVE
COLOR
VIS
GRID
SNAP
STATIJS
WIRE
3V
HELP
Main
MINE
MACIllNES
ANALYSIS
SE11JPS
LI
L
~
B P F
"l
R [7
4 ~
L>.
CfR
v
.25x .Sx 2x
WIN EXT PRE
L>.
H
v
DOL
CAM PER
Fig. 1. The default screen layout of CAP with a solid model of a battery powered scoop displayed.
main menu area of the program. This is a changeable
menu several levels deep that controls access to the
majority of the functions of the program. From this
menu you can open and save files, input mines,
machines, and lamps, edit and delete objects, run the
analyses and change the program settings.
The upper left portion of the screen contains a user
defined menu. If the user determines that there are
functions in the main menu that they use more often
than most, the controls for those functions can be
permanently displayed here so they are readily
available.
Prompts to the user and other pertinent information
is displayed in the banner across the top of the screen.
Information about what the program is currently doing
is generally displayed here. Input into the program is
often made through dialog boxes, such as the one
presented in figure 2. This particular dialog box
allows the user to position a lamp, change its type or
tum it on or off.
The models of machines and lamps used by CAP
are usually created with AutoCAD. The AutoCAD
DXF file is translated to the CAP file format with a
utility program supplied with CAP. However, simple
machine models can be created by using either a text
editor or a separate model build utility supplied with
CAP. Actual mine layouts can be input to CAP by
scanning existing mine maps and vectorizing the
resultant image into AutoCAD. Mines can also be
created from scratch in AutoCAD.
CAP's Visibility Analysis
Visibility is a significant problem in the underground
mining environment. During the 1980's, the Bureau of
Mines sponsored research to determine minimum
visibility requirements for three classes of underground
mining machines: shuttle cars, scoops, and continuous
miners. Using structured interviews and on site task
analyses involving approximately 100 subjects, the
researchers first identified the tasks involved in the
operation of each class of vehicle, such as loading,
hauling or unloading. Then, the machine operators were
interviewed to identify the visual information required to
perform each of these tasks. For example, a shuttle car
operator performing a loading task would need to see
the positioning of the shuttle car under the tail boom of
the continuous miner. While hauling, the location of the
shuttle car and any obstacles in the roadway would be
required. Following the interviews, the operators were
observed while performing the tasks to verify the visual
information requirements established in the interviews
(Sanders and Kelly, 1981).
Once the requirements were determined, a
methodology was developed to identify specific points
in the Front-to-Back, Side-to-Side, and Vertical planes
that must be visible to the operator to satisfy the visual
requirements. These points, called Visual Attention
Locations (V ALs), were defined in reference to generic
locations on the machine. This allowed the V ALs to
3
15:15
Lamp Data
XLoc. ALPHA STATUS
1
220
.
000
1
~
ON
YLoc. THETA TYPE
1 42.000 1
~
MCJ2200
ZLoc. BETA Name
1 37.500 1
0.000 I Lampl
Fig. 2. The screen layout switches to three view mode when objects are being positioned.
The dialog box allows the input of precise information for the objects.
apply to all configurations of a particular equipment
class. For instance, while tramming, a shuttle car
operator must be able to spot an obstruction on the
ground while there is enough time to stop the vehicle.
The location of one of the V ALs associated with this
requirement is described in Table 1.
Using this methodology, the procedure involved in
computing the location of the V AL is the same even if
the length of the equipment, the operator's position, or
the height of the equipment is modified. Therefore,
the VAL is applicable to all equipment in the shuttle
car class.
The results of this VAL research have been
incorporated into the CAP package. The CAP
visibility model automates the task of determining
whether or not the required V ALs are visible to a
Coordinate Plane
Front-to-Back
Side-to-Side
Vertical
selected human operator, currently either a fifth
percentile female or ninety-fifth percentile male. The
output is a relative visibility rating for the machine
which can be compared with results of alternative
machine designs. CAP also provides both graphical
(figures 3 and 4) and tabular output to pinpoint any
machine parts which obstruct visibility.
CAP's lllumination and Glare Analysis
Due to the perpetual darkness of the underground
environment, illumination and glare are factors which
must be considered when designing equipment for
optimal visibility. Federal regulations specify that
certain surfaces within a miner's normal field of vision
must be illuminated to 0.06 footlamberts while
Position of VAL
Front Edge of machine +
Necessary Stopping Distance
Machine Center Line
Floor
Table 1. Location of a VAL associated with tramming a shuttle car
4
10:40 VIEW: REAR RIGlIT ZOOM: 0.356 Cur Mach: 102HP SNAP WV HIDE
JLLUM
VIEW
PRINT
CLRJLL
SHOW
SHADOW
ll.AMP
ELAMP
VPANEL
HPANEL
EPANEL
CLRALL
OPEN
SAVE
COLOR
VIS
GRID
SNAP
STAruS
WIRE
3V
HELP
Machines
INPUT
EDIT
DELEIE
VISIINVIS
ALL VIS
ALLINVIS
VALINVIS
SIllELDS
LAMPS
PANELS
Ll
L b,.
B P F
"l
R
l7
4
A
CIR
v
.25x .5x 2x
WIN EXT PRE
H
A
DOL
v
CAM PER
Fig. 3. A shuttle car and continuous miner in a entry. The pillars are from a scanned mine map.
self-propelled mining equipment is being operated.
The 0.06 footlambert level is a measure of luminance,
or photometric brightness. It is a product of the level
of illumination (incident light) impinging on a surface,
and the reflectivity of the surface. Bureau research
has shown that the 0.06 footlambert level is adequate
for most mining tasks and is low enough so that
operators will not experience severe adaptation
problems when moving from illuminated to non-
illuminated areas of the mine (Lewis, 1986).
Unfortunately, in attempting to meet these
illumination standards, mine equipment designers
sometimes aggravate another mine lighting problem -
glare. There are currently no standards related to
glare in underground mining. The regulations state
only that designers should attempt to minimize glare
when developing machine illumination systems.
Obviously, performing all of the calculations required
to compute glare for a multitude of machine and lamp
types, with the possible combinations running into the
thousands, is too tedious and costly to be practical
using manual methods. The result is that the designer
is impeded significantly in solving for an optimal
illumination system that minimizes glare.
The CAP illumination model eliminates these
problems by turning the computational portion of the
lighting design task over to the computer. The
software allows the lighting designer to concentrate on
adjusting the configuration of the illumination system
to minimize the potential for glare while still providing
enough illumination to conform to the Federal
regulations.
To perform an illumination analysis using CAP, the
user lays out a lighting system on a machine model.
Each lamp has associated with it an illumination profile,
previously measured along selected angular orientations
from the lens of the lamp. Using these profiles, the
software can calculate the light output at any location in
the scene. The model does not take into account
reflected light, but it can correct for the shadowing
effects of machines or other objects in the scene.
The illumination profile model was initially based on
the inverse square law of lighting, where illumination
varies inversely as the square of the distance. The
method of gathering data consisted of recording a single
measurement of distance at a illuminance reading of
2.00 footcandles (fc) for each vector in 5 degree
increments around the luminaire. This data was then
used to find the illumination at any other distance along
a given vector using the inverse square law.
There were some inherent problems with this
method. By basing the illumination calculations on a
single measurement along the vector from the luminaire
to the measurement point, relatively large errors could
occur through inaccurate data collection. In addition,
the inverse square law assumes that a luminaire is a
single point source of light, a condition that clearly
cannot be achieved in the real world. This discrepancy
5
10:41 VIEW: ........ ZOOM: 0.356 Cur Mach: 102HP SNAP WV HIDE
IllUM Machines
VIEW
INPUT
PRINT EDIT
CLRIll DELETE
SHOW VlS/lNVJS
SHADOW ALL VIS
ILAMP ALLINVIS
!!LAMP VALINVIS
VPANEL SHIELDS
HPANEL LAMPS
EPANEL PANELS
-
-
CLRALL
n
OPEN
::l. g- O
[]t1P
[][]
I
lSf
01
[] SAVE .,u
an
0 []ffj
~ l L$
e;;;;;= ::::::...
~ I
.d" I T
~
~
l:r
~
rn----
- - - ~
to ro
COLOR ~
.Ll
L
~
VIS B P F
GRID
"I
R [7
SNAP
H
.0.
erR
V
STA1US .2Sx .5x 2x
WIRE
\
WIN EXT PRE
3V
4 ~
.0.
DOL
V
HELP CAM ORT
Fig. 4. View from inside the shuttle car cab. The V ALs are represented as blocks.
causes calculation errors to occur within a distance of
approximately five times the longest dimension of the
luminaire. The method currently used for modelling
illumination avoids both of these potential problems.
Seven measurements of distance and illuminance taken
along each vector at 10 degree increments are used to
develop a best fit equation for each vector. This
equation is then used to interpolate the illumination at
any point along that vector.
In more detail, the measured distance data is
transformed by raising the distance value to a
variable exponent and then taking the reciprocal. This
transforms it into (approximately) linear data. The
footcandle measurements are transformed by a base 10
logarithm. This reduces the scale and the "weighting"
effect found without this method. Without taking the
logarithm first, the scale of absolute footcandles is
very large. The regression naturally minimizes the
absolute distance of the fit line, and has the effect of
giving the same quantity of error across all
measurements. The problem with this is that the
percentage of error can become quite large at the
greater distance (and therefore smaller fc)
measurements. Thus, the logarithm has the effect of
reducing the footcandle scale, and evening out the
percentage errors.
A regression analysis is performed on the
transformed data with a "best-fit" line to determine the
linearity of the transformed data. The R-squared
value or correlation coefficient of the regression
equation is evaluated to determine precisely how linear
the data is, 'and therefore, how close the fit is to the
actual data. (A preliminary R-squared value of 0.97 is
presently used in the calculations.) If the fit is not close
enough, the exponent is modified and the process is
repeated. This continues until an equation is found that
best fits the data.
This final equation for each vector of a given
luminaire is (by definition of the process) a good fit to
the way illumination decreases as distance from the
luminaire increases. Calculated values of illuminance
from this equation, according to a statistical analysis of
preliminary data. has a mean very close to D% and a
standard deviation of between 3 and 10 percent from
measured values of illuminance. This indicates that
more than 95% of all illumination values calculated
using this method are within 20% of the actual values
and are evenly distributed. Consequently, with multiple
lights there is a higher probability of balancing the error
factor. Using a regression analysis of the data and
allowing each variable in the calculation of illuminance
to be derived excludes potentially errant assumptions,
such as assuming the luminaire is a point source of
light. Hence, calculations of predicted illuminance are
made as "clean" as possible.
Further, the regression analysis method enhances
error checking in that seven measurements of
illuminance and distance are made for each selected
6
Distance Measured Inverse Sq. Law Regression Eq.
(meters) (fc) (fc) (fc)
0.305 79.30 118.8 82.40
0.914 12.30 13.2 11.73
1.684 3.82 3.9 3.72
2.352 2.00 2.00 2.00
4.124 .66 0.65 0.67
5.954 .32 0.31 0.32
7.784 .18 0.18 0.18
Table 2. Comparison of illuminance vs. distance for measured and computed values
vector. If an error should occur in either the
measurement process or in transcribing the data, it
will readily show up in the calculation of the
regression, given the sensitivity of the calculation.
M ~ s t a k e s will show up clearly as absurd results. It is
also possible to get a: statistical analysis of the
appropriateness of the "fitted" curves over the entire
scale of the light, and so determine the quality of the
measurements and analyses performed with that light.
Therefore, the regression method enhances the
monitoring of human error that would normally occur in
the process of data collection and input, as well as the
analysis.
As an example, the Table 2 presents a comparison of
illuminance versus distance for measured and computed
values.
In CAP, illumination measurement panels represent
the areas (roof, ribs, floor and face) that must be
illuminated to 2.0 footcandles to meet the Federal
regulations. These panels can be sized and positioned
10:26 VIEW: ***+* ZOOM: 1.953 Cur Mach: 102HP SNAP HN
Fig. S. A continuous mining machine with three of its lamps turned on. There are four illumination
measurement panels, one on the floor, one at the face and two at the sides.
7
10:36 X=366.000 Y=60.000 MODE: INFO ZOOM: 1.000 Cur Panel: FLOOR SNAP HIDE
Point selected: 10.95 PC, Block average: 6.87 Pc.
n.LUM
m""""
I I I
I
I

I
I
I
VIEW AJpha= 0.000
FILE: 102HP
PRINf X=90.000
CLRn.L

Y=!08.000
SHOW
0<2.0fc
QCK UN DEP
SHADOW
1 I' , I'
!LAMP UNP T1LE
ELAMP
1 I' , I'
VPANE!. PS HP COL
HPANE!.
r-J C C
r
BUT sa INF
EPANE!.
0
11NC RINC
CLRAll
mI
UlG AXS UB
OPEN.
[JJ
DONE
SAVE
F
Q
0
Q
\.
COLOR
Ll
L
b..
VIS B P F
GRID
"'l
R [7
SNAP
I
I
I
I
I
I I
I I I
I
I
H
.0.
erR
V
STAlUS .25x .sx 2x
WIRE WIN EXT PRE
3V
H
.0.
DOL
v
HELP CAM PER
Fig. 6. The illumination measurement panels are unfolded to display the results.
The crosshatched blocks indicate an average illumination of at least 2.0 footcandles.
around a machine model (Figure 5). The output of
the illumination analysis can be displayed on these
panels so that quick comparisons can be made
between lamp layouts (Figure 6).
Using the calculated values of illumination, the
user may perform Disability Glare comparisons by
calculating the luminance contrast of objects of
interest along selected lines of sight. The luminance
contrast is the relationship of the luminance of an
object and its immediate background. In an
underground coal mine, the background is usually the
walls, floor or roof of the tunnel being mined, which
has a known luminance value.
Hardware
IBM compatible 386/486 computer
Math co-processor
8 Mb of RAM (16 is preferred)
At least 10Mb Hard Disk space
VGA monitor (Super VGA preferred)
256 color video card
Microsoft compatible mouse.
Future Plans for CAP
The CAP package is currently undergoing BETA
tests with an expected release of Version 1.0 in late
1993. The hardware requirements for CAP are listed in
Table 3. AutoCAD Release 11 or 12 is also necessary
to generate complex machine models. However, CAP is
supplied with a library of generic mining machines that
suit most needs. There is a charge to cover the
Bureau's licensing fees for the graphics toolkits used by
the software.
There are many advantages to using computers to
assist in the design of mining equipment. Computers
permit quick evaluation and comparison of different
Software
DOS 5.0 or greater
AutoCAD Release 11 or 12 (optional)
Table 3. Hardware and software requirements for CAP
systems, such as when trying to optimally position
lamps. Infonnation that is difficult or tedious to
incorporate into a design with manual methods, such
. as anthropometric data, can be conveniently accessed.
Graphical feedback not only enables the designer to
quickly identify potential problems, but also enhances
his or her ability to communicate ideas to the client,
increasing the likelihood of acceptance of novel
designs.
8
In the future, the Bureau hopes to enhance CAP
with additional capabilities. Work is already
underway to automatically generate the path network
composed of the entries and crosscuts of the mine
model. This will allow computerized timing studies to
be perfonned, such as the time needed for personnel
to escape from a section. Operator reach and
accommodation models may also be added. These
programs would help the designer to determine
whether operators from a selected population would be
able to fit into the crewstation and reach the controls.
As with the illumination and visibility models, the
unique requirements of the underground mine worker
would be taken into account, induding provisions for
miners wearing hardhats, metatarsal boots, and self
contained breathing apparatus.
Acknowledgements
A software project such as CAP requires the
expertise of many professionals. The author wishes to
Acknowledge the contributions of USBM employees
Sean Gallagher, Audrey Glowacki, Chris Hamrick, Alan
Mayton, James Rider, E. W. Rossi, Paul Sonier and Tris
Betts during the development of CAP.
References
Lewis, W. H., 1986, Underground Coal Mine Lighting
Handbook, in two parts: Background, U.S. Bureau
of Mines Infonnation Circular 9073, 42 pp.,
Application, U.S. Bureau of Mines Infonnation
Circular 9074, 89 pp.
Meister, D., 1971, Human Factors: Theorv and
Practice, Wiley Interscience, 415 pp.
Sanders, M. S., and Kelley, G. R., 1981, Visual
Attention Locations for Operating Continuous
Miners, Shuttle Cars and Scoops, Final Report, U. S.
Bureatl of Mines, Contract J0387213, 132 pp.