Note: Descriptions are shown in the official language in which they were submitted.
CA 02212742 1997-08-11
W O 97/22096 PCT~US96/19664
PAOEID llERATIVE DECISION TRAINING SYSTEM AND MEI ~OD
Technical Field
This invention relates generally to training systems and m~thodc More
particularly, this invention relates to ~l~c~ti~n~l andfor voc~tion~l training systems and
mPthorl5 for improving the paced iterative decision-making ability of person who is lc~Luilcd
to make quick a~ ces~ of r.h~n~ing events. Although the invention has broad utility, it
has proven e~e~i;ally well suited for use in the context of a lumber grading/sorting training
system and method.
~ackgrolmd Art
]L0 Many occup~tions and activities require quick and decisive evaluations of
given events or a series of given events. Unfortunately, before a worker can cQn~ict~ntly
make the proper evaluations of such events, the worker must have experience. However,
tr~-litir,n~l training systems and meth~lc for providing the nPcec.c~ry experience are
,lle~ live and/or expensive. These tr;~dition~l systems and methods may include text
'15 books, work books, pop quizzes, written ~x~r~ ;Qnc, oral c~ .";~ ;onc classples~ "iQnc, hands-on d~."~ Lions, lectures, flash cards, field trips, factory tours, and
lalx,l~oly ~ ,elullents. While these are valuable and useful pe~ ogic tools, they are
in.c~lffirient when it comes to developing a person's skill for making quick and accurate
de~icionc
Tr~-1iti~ n~lly, the best methods and systems for developing these skills are
analogous to te~-~hing a student to swim by throwing the student into the water at the deep
end. A student is placed into the role of a decision-maker within an actual workem,ilulullcll~ The student's de~ icionc are reviewed by an instructor who corrects the
student's mict~k~s The student learns from the student's mict~k~c so that the student may
~5 perform better next time.
Disadvantages of such traditional training methods and systems is the
expense of training a student in an opeldling work environment. It is expensive because the
student is likely to make many errors that must be corrected. This activity reduces
productivity and may lead to unnecessary safety and quality concerns. A training system
and method is needed in which a student has the opportunity to learn and gain valuable
CA 02212742 1997-08-ll
W 097/22096 PCTAUS96/19664
e~,~iellce without working in an actual work environment and unn~cee~rily risking safety
and quality.
Fx;1~..p'~.s of occup~tion~ and activities that require a quick and honed
d~;c;on-making ability include a fish counter and a lumber grader. A fish counter identifi~
and counts the various species of fish that pass through a dam. A lumber grader de~ s
the quality or grade of a piece of lumber as it travels past the grader at a grading station.
Lumber is sorted based on the quality or grade dt;Le""."ed by the grader
upon visual inspection of the lumber as it travels past a grading station. The price at which
the lumber may be sold is at least in part based on the grade ~cii~ed by the grader.
Grading lumber takes a high degree of skill and experience. Unfortunately, existing ways to
train a novice grader are costly and time con~llming
Lurnber grading systems exist that help improve the quality of lumber
grading by rotating the piece of lumber so that the grader can see all sides of it before
giving it a grade (such as U.S. Patent No. 3,759,384 to Holmberg et al.). However, no
cost-effective training system exists to help new and existing graders gain valuable
experience and thus, improve the overall quality of lumber grading. Since the quality of
lumber grading is Illtim~t~.ly dependent upon the skill ofthe grader, a cost-effective training
system and method is needed to help train new graders and hone the skill of experienced
graders.
Disclosure of the Invention
The invented training system and method solves the ~ iul~ed
problems by introduction of a ~im~ tion of a series of events for a person to grade and a
means of l~col~ g the grade of each event in the series for COIllp~hiSOll to one or more
person's grades of the same series of events. In the invented training system and method, a
grader is presented a series of paced events for the grader to judge. The graderls recorded
~es~",~,.l of the events are compared to a separate recorded ~c.c~ of the same
events by the same or di~llL grader. The separate recorded s~s~e~ "l may be created
by an master grader or by the same grader. A purpose of the training system and method is
to provide a ~im~ ted envilul"llt;nL in which a grader's decision-making skills can be tested
and improved. Typically, expert feedback is utilized to improve the novice grader's ability
to assess, evaluate, judge or grade each event in a series events. A com~isoll of the
I
CA 02212742 1997-08-11
W O 97/22096 PCTnUS96/19664
novice grader's 5~ C~ h.ll to a master grader's ~cs.occm~nt of the same series of events
provides the expert feeclb~ek needed for the ~p,.,pliate pe~ c effect.
The invented training system and method has a potentially wide arplir~tif n
It is p~re,~bly ~l~ci~P~ to help train lumber graders in a lumber sorting system. The
training system and method reduces the time and expense required to train a novice grader.
The training system and method records a video of traveling pieces of lumber on a
conveyor system using a camera and video tape recorder/player (Vl~/P), which may be
coupled to a monitor. An event detector detects each piece of lumber so that one piece of
lumber is graded per event. A comruter controller determines the timing of each event
based on ~yncl~ ion (frame) signals from the VTR/P and an event signal from the
event detector. The resulting video and recorded event il~l",~Lion is an event-cim~ tion
(or simply a sim~ tion).
Plt;r~l~l)ly, the cimlll~tion is shown to a master grader in ~csoei~tion with a
comrlter controller. The master grader sits at a grade entry station ~GES) and grades the
pieces of lumber that pass by on the video monitor. Each event is graded by the macter
grader and each event-associated grade is stored to produce a grade dataset. Thecim~ tit~n and the ~cco~qted dataset is a training program.
To learn how properly to grade or judge lumber, a trainee sits at a grade
entry station while a training program is replayed. The trainee grades each piece of lumber
as it passes through the video frarne. Later, the trainee's grading is colllpaled to the master
grader's baseline grading. This process can be repeated with the same or difIèl~llL training
programs as many times as desired. The net result of the system and method is a faster, less
e~pensive method of training a trainee than previously available.
These and other advantages and objects of the present invention will be
more readily understood after a con.cidçration of the dra~vings and the detailed descri~.Lio
ofthe pl~r~ ;d embodiment which follows.
. BriçfDescription ofthe Drawin~s
Fig. 1 is a general schematic bloclc diagram a training system, constructed in
accoldallcç with a preferred embodiment.
Fig. 2 is a s~ ,c diagram of the training system of Fig. 1 showing the
colllponellL~ usçd to record a video of a series of events.
CA 02212742 1997-08-11
W O 97/22096 PCTnJS96/19664
Fig. 3 is a sch~m~tic diagram of the training system of Fig. 1 showing the
co..,l,onc;l,L~ used to present a video of a series of events and record the grader's evaluation
of each event.
Fig. 4 is a fiow chart of a training method, in acccl~lallc,e with a pl~ert;ll~d5 embo~lim~ntq
Detailed Description of the Drawin~s. Preferred Embodiment. and
Best Mode for Carryin~ Out the Invention
~ Itho~lgh the preferred embodiment for the invention is described in the
context of a lumber grading training system and method, those who are skilled in the art
10 will understand and appreciate that the invention has a potentially wide applic~ti~n to any
e~ ~tion~l and/or vocational training system or method. The training system, in
accc,lda~lce with a ple:rt;llc;d embodiment, is generally shown at 10 in E~ig. 1. In general, the
training system inr.l~ldes an event-simulation subsystem 12, a plrsf~ ;on-and-entry
s~ y~Lelll 14 and a controller 16.
Event-~im~ tic n subsystem 12 creates, produces and/or records any
se ~ once of events (i.e., series of events) for playback later to a user of prÇs~nt~ti~ n-and-
entry sul,~y~ ln 14. An l'event" is any action or thing that occurs during a p~li~;ular
interval of time and a "series of events" is any sequ~n~e or succession of events. For
~ l le, a series of events may include pieces of lumber on a conveyance system, fish
20 passing through a dam, etc. An event has characteristics (i.e., properties, q~ itie~ etc.)
that are capable of being objectively and/or subjectively determined by a person. A person's
evaluation of the cll~ ile-i~lics of a piece of lumber is the piece's "grade". For fish passing
through a dam, a person may quickly count the a,vplu~illlale number of fish of specific
species passing through during a short time interval. Preferably, the ~im~ tion created by
25 event-sim~ tion subsystem 12 is a video leco,dillg of an actual series of events.
Alternatively, the simlll~tion may be an animation (hand-drawn or computer-g~l.el~Led)~
slide show, virtual reality, etc. The event-~im~ tion subsystem p,e~el~ly includes a video
carnera, a VT~/P, a video monitor, and an event detector.
Pl~s~ )n-and-entry subsystem 14 receives the sim--l~tion from the
event-~im~ ti~n~ubsy~L~lll and ~.~s~ it to the user. The user views the ~imlll~tiQn and
CA 02212742 1997-08-11
W 097/22096 PCTnUS96/I9664
enters a grade for each event. The p~sel~ ion-and-entry S~y~ pl~;rc;l~ly includes a
video monitor, a VI~IP, and a grade entry station.
Controller 16 coordinates and controls the event-sim~ tion sub~y~lem and
the pr~;c~ ion and-entry sul)~y~ell~. The controller tlictin~lich~os between individual
events based on a signal from an event detector. It also ~ccori~tes and stores the user's
grades with the proper events. After two or more ev~ tionc of the same cimlll~tinn, the
cul~Ll~lle. ccm~ s the ev~ tinnc and reports the results. Plt;re~ly, the cc,llL,olle,
inr1~ldes a computer (with a processor and memory/storage systems) and a display.
Figs. 2 and 3 show the preferred embodiment of the training system in more
] 0 detail than shown in Fig. 1. The preferred embodiment of the training system and method
ope,~Les in two basic modes: production mode and grading mode.
Fig. 2 shows the training system in the production mode. When operated in
the pr~clucti~ n mode, the preferred embodiment of the invention includes a high rocollltinn
video camera 20 and a video tape recordel/~,ldyel (Vl~/P 22) which collectively record
] S video images of pieces 21 of lumber on a conveyer system 23 as they travel seriatim past a
grader 30. A video monitor 24 may be ~tt~ch~d to the ~IR/P. In Fig. 2, monitor 24
shows an ."~l.,)lo of one frame of the video tape being recorded on the VTRIP. The
mûl~lLor shows pieces 21 of lumber on conveyer system 23. While taping, the Vl~/P sends
frame .l~lllla~ion to compllter 26. The comrutçr may have display 27 ~tt~rh~d thereto.
~0 In the preferred embodiment, no video images are stored on the computer.
It is n~cç~ .y- to determine when each event takes places so that the events
may be graded individually. In the preferred embodiment, each event is piece 21 of lumber
travelling seriatim (i.e., one-after-another) by conveyance system 23. An event detector 28
(pl~l~bly, in the form of a progl~ --able logic circuit (PLC)) detects and dt;l~l"l,l~es
when a new piece of lumber may be graded and sends an event signal to computer 26.
Event detector 28 is used to provide grade zone timing definition (an event signal~ to the
C comI ~~t~r 26. The event timing signal is to be provided from the conveyance system 23
controller (typically a discrete output from a PLC) which signals when a new piece of
lumber has been presented for grade inspection to the grader 30. The pairing of the event
30 timing signal for grade entry coupled with the VTR/E' fiame i"ro...~ ;on provides
~iy~cl~ it)n of the board video image and its associated dataset which is stored by the
CA 02212742 1997-08-11
W O 97/22096 PCT~US96/19664
c mrut~r 26. Based on the event signal sent by PLC 28 and the frame il~lll~Lion sent by
VI~/P 22, c~-mp~ltPr 26 detPrmines when each event occurs. The frame and event
illru~ ;on is stored in the comrut~rs primary and/or secondary memory systems. For
~ lle, frames 90-150 may be a first event, frames 151-209 may be a second event,frames 210-275 may be a third event, and so forth.
The resulting co~ aLion of the video tape of the series of events and the
frame and event il~ll,lalion stored on the computer is called a ~im~ tion program. A
cim~ tion program may last any given length of time. Preferably, a sixty to ninety minute
program is used to he1p train lumber graders.
Fig. 2 also shows a grader 30 using grade entry station (GES) 32 or any
other data entry station. The grader and GES is not necess~ry for the production mode;
however, the production mode and grading mode may occur concurrently. Thus, the
grader may view and grade the pieces of lumber on the conveyer system while a ~im~ tion
program is being produced using the same pieces of lumber. The grader's grade may be
~c~oc;~tecl with each event as it occurs.
Fig. 3 shows the training system in the grading mode. Grader 30 using
grade entry station 32, views a ~imlll~tic)n program on video monitor 24. The video on the
video monitor is produced by Vl~/P 22. The V~/P sends a signal to comr~-t~r 26 that
include frame il~llll~ion on the ~im~ ti-~n prograrn being viewed by the grader.Each event (i.e., pieces of lumber) is presented to the grader at paced,
defined time intervals. This means that the time intervals have a defined ~i.e., pre-selected)
length and are s~pal~le from each other. There should be little or no overlap between time
intervals for events. In the preferred embodiment, the interval of time for each event is
generally the same. Typically, the time interval in which each piece of lumber is pl~sellLed
to the grader for grading will vary according to the speed c~pahilities of the conveyance
system as well as the desired visual inspection grading rates. In lumber grading a~plic~l;Qm
this timing interval is typically between .5 to 1.0 seconds for each board event. Also, the
GES sends a signal to the computer indicating the grade d~ign~te~ and entered by the
grader. Based on the frame il~,l,ldlion sent by the Vl~/P and the stored associated frame
and event i~ ion, the computer associates the grader's grade with each event of the
~im~ tion program.
I
CA 02212742 1997-08-11
WO 97/22096 PCT/US96/19664
Preferably, the comp~tçr is an mM-Go.1.r~ le comruter with a primary
memory system (i.e., RAM) and secondary memory systems (e.g., floppy and hard disk
drives). The ccmruter also p~t;;r~ly includes a KIETH~EY METRABYIE 9~ port VO
(input/output) card. Grade entry station 32 and P~C 28 are conne~ted to the comruter via
S this I/O card.
In 1umber sorting systems, there are two common means of entering a
grade for a piece of lumber. The first means is depicted in Fig. 3 and to some extent in Fig.
2. The grader mans a GES and pushes buttons that indicate the grade of the piece of
lumber that the grader is viewing. The GES is typically a keyboard similar the GES 32
10 shown in Figs. 2 and 3. The other commcn grade-entry means is where the grader
personally turns over each piece of lumber to inspect it closely. Once the grader d~Le~ es
the piece's grade, the grader writes the grade on the piece itself. The grade is optically
scanned into a comrl~ttor so that the piece may be properly sorted
The ~ f;l-ed embodiment may utilize the both co~lnlon grade-entry
]!5 means. The first has been tii~c~ ed above. Alternatively, the second col".1~ol1 grade entry
system of personal inspection and hand writing the grade may be utilized. The grader
enters the grade of each piece by using a tablet entry system. GES 32 would include a
tablet entry system on which the grader's writes symbols in(liç~ting specific grades. The
GES and/or the com~n~ter translates the grader's written symbols into grades that can be
20 stored in the dataset.
When the training system is in the grading mode, a grader views a program
and grades each event to produce a dataset. Preferably, the dataset includes the grader's
grade of each event in the program. Once at least two datasets are created for a program,
they may be co",~ d. The computer compares the datasets and the results are stored in
25 the computer and ,~lt;ft;l~ly displayed on display 27.
The co~ aled datasets may be created by the sarne grader or by difre,~"L
graders. When training a novice grader (i.e., trainee), it is desirable to use a master grader
to create a master or baseline dataset for the simulation program. The trainee can learn
from the trainee's mi~t~kçs by viewing a report on the differences between the trainee's
30 dataset and the master dataset. From such comparison, the trainee can learn how to grade
better by c ~ g how his evaluation of the events COIl~ ;S to a master grader's
i
CA 02212742 1997-08-ll
W O 97122096 PCTnJS96/19664
eva~uation of the same events. Also, this system may be used by a single grader multiple
times. A grader may improve ~ y in grading by comparing the grader's evaluation
ofthe same events.
In Fig. 4, the preferred embodiment of the invented t~aining method is
shown. The method starts at 50. At 52, the grader is first presented a series of events. At
54, the grader first evaluates the characteristics of each event. At 56, the grader first enters
such evs~ ti~n~ At 58, the grader's ev~ tion~ of each event of the series are first stored,
thereby producing a first dataset. The resulting first dataset is shown at 59.
At 60, the training method determines whether a second dataset exists. If
so, the first and second datasets are co~ a,ed at 70. Otherwise, a second dataset must be
created; therefore, at 62, a second presentation of a series of events is given to the same or
di~ele." grader. At 64, the grader does a second evaluation of the characteristics of each
event of the same series of events. At 66, the grader does a second entering of such
evaluations. At 68, the grader's evaluations are stored to produce a second dataset. The
second dataset is shown at 69. At 60, the answer to the whether a second dataset exist
must be "yes" because a second dataset was created in steps 62-68. Next at 70, the first
and second datasets are co~ )~ed. At 72, inro.llla~ion based on the results based on such
COIII~ 1C is reported and such report may include storing the i~-r~",.~ ;cn for later
retrieval. The training method ends at 74.
Of course, those who are skilled in the art understand and ap~ ,;aLe that
the des~rirti~ n above describes the preferred embodiment and the drawing and does not
limit the spirit and scope of the invention. Other possible alternative embotlim~nt~ exists.
For example, rather can video taping actual events to produce the cim~ tiQn, a computer
can create a virtual reality (VR) envi,~ llwlL. Using a VR scenario, the user would don a
VR helmet and glove. The user would see computer-generated three-rlim~ncion~l images
of events (such as pieces of lumber) and the GES. A computer would deL~ ine which
virtual button the user was pressing to indicate the grade of the event. The VR scenario is
only given as an example, but other possible modifications may be made without departing
from the soul and breadth of the invention.
CA 02212742 1997-08-11
W O 97/22096 PCTAJS96/19664
Industrial Applicability
The invented training method and system now may be understood to
provide a method and system for training a grader to develop and improve the grader's
ability to make decisions and ev~ln~ti~ nc under the pressure of time. Those who are skilled
in the art will understand and appreciate that the invention has a potentially wide applic~ n
to any e l~lc~tional and/or vor~tiQn~l training system or method.
Fx~mrlçs of training scenarios in which the invention is an appl~,p-ia~e and
cost-effective training system and method include: grading pieces of lumber; grading pieces
of wood veneer; counting fish passing through a darn; grading-the quality of finished or
O processed goods on an a~ellll,ly line; grading the quality of raw products before entering a
processing/assembly line; e,~ g x-rays of luggage at securitv check points; t;Y .~
medical test results; reviewing plays from sporting activities to judge the proper foul or
penalty to call; reviewing plays in football games to determine the proper of~ensive or
defensive audible; and any perception/observation testing.
While the p-'~;r~ d embodiment and best mode of the invention have been
close-l variations and changes may be made without departing from the spirit and scope
ofthe invention
-