Barcode tracking of Rolling Stock
Steven Dorst
Hello,
Our club is beginning to investigate the use of bar code tracking of
rolling stock on our layout. We have 1800+ cars and 700+ engines that
belong to members currently certified to run on our railroad. During
our 4 hour public operating sessions, we run with 7 mainline cabs, and
have 30-50 trains seen during a session.
We envision putting a bar code on the bottom of every piece of rolling
stock, then putting multiple readers at various critical points on our
layout.
I'd like to know if anyone else out in r.m.r land has experience in
using bar code technology in conjuction with model railroading.
Thanks in advance,
Steven J Dorst <do...@usa.net>
HO Scale Trainmaster
East Bay Model Engineers Society at the
Golden State Model Railroad Museum
Visit us on the web at:
http://www.hooked.net/users/robertjm/gsmrm.html
Daniel A. Mickey
Steven Dorst wrote:
>
> Hello,
> Our club is beginning to investigate the use of bar code tracking of
> rolling stock on our layout...
snip
> We envision putting a bar code on the bottom of every piece of rolling
> stock, then putting multiple readers at various critical points on our
> layout.
snip
How much are bar-code readers?
I can anticipate a wonderful new world of car waybills and traffic
generator software, with the cars identified going into or out of
various yards and spurs. The software engineers would be in
computer-heaven! You could even tick off brownies for crews who put
cars in the wrong places.
Seriously, if the costs were low enough and the system easy-enough to
understand, there could be some good developments from this.
Dan Mickey
Fritz Milhaupt
Daniel A. Mickey wrote:
[snip]
> How much are bar-code readers?
>
> I can anticipate a wonderful new world of car waybills and traffic
> generator software, with the cars identified going into or out of
> various yards and spurs. The software engineers would be in
> computer-heaven! You could even tick off brownies for crews who put
> cars in the wrong places.
The last time I checked, you could get a handheld CCD bar code reader
(like the kind used in a lot of video stores) for about $150 in "wedge"
form, where you would connect it into a keyboard port.
What I'd envisioned was not barcoding the cars as much as using a system
where each car has a clear PVC envelope just large enough to fit a 3" by
5" card with the move information into. The envelope would be labelled
with the car's ID information, and move with the card.
Each time a move is generated, it would be assigned a serial number,
which would be printed in barcode form on the card, along with the usual
human-readable move instructions (Car ID, origin, destination, lading,
routing, etc.). When a crew finishes its work, the sum total of the
paperwork required would be to pull the move cards from the car
envelopes, then log the moves as completed by passing the car cards
underneath the scanner, so that the car movement program could update
its car location records. The envelope would remain in a pocket on the
layout fascia, adjacent to the location at which the car is spotted.
In fact, under this system, the envelope might not even be necessary.
-fm
Preserving the memory of the Pere Marquette- one HO loco and car at a
time.
To reply via e-mail, remove "lma." from the return address shown above.
MSTie #78116 (Better late than never!)
John Purbrick
In article <336901...@lma.rust.net>,
Fritz Milhaupt <milh...@lma.rust.net> wrote:
>Daniel A. Mickey wrote:
>[snip]
>> How much are bar-code readers?
>>
>> I can anticipate a wonderful new world of car waybills and traffic
>> generator software, with the cars identified going into or out of
>> various yards and spurs. The software engineers would be in
>> computer-heaven! You could even tick off brownies for crews who put
>> cars in the wrong places.
>
>The last time I checked, you could get a handheld CCD bar code reader
>(like the kind used in a lot of video stores) for about $150 in "wedge"
>form, where you would connect it into a keyboard port.
>
>What I'd envisioned was not barcoding the cars as much as using a system
Much better to barcode the cars if at all possible. Then each car can be
tracked from one end of its journey to the other, and as the earlier poster
said, you can assign a score to the operator, plus so many points for on-time
delivery and minus so many for being late or getting a shipment in the wrong
place. An operator's switch list would be in the form of a printout saying
which cars are to go where.
But $150 a reader is a bit pricy for most of us. You can get the bare interface
component from Hewlett-Packard for about $25, and this gives an electronic
output which a microcontroller could turn into a serial data packet to send
to a computer; I'd bet that one micro could service several (many?) readers.
Is the motion of a model railroad car regular enough to get a good scan
most of the time? That's a question.
John Purbrick
David Gibbons
I ordered a catalog from a company who sold all sorts of bar code
readers, and they are expensive - the ones that can read from a distance
with a laser run $400 to $800 each.
I suspect those of us who are interested in this will need to develop a
reasonably priced approach on out own. The thing that may help is that
we can control the relationship between the scanner and tags pretty
well.
The other challenge is getting a system which can read tags at the speed
that our model trains run at.
Should tags run parallel to the length of the car. or a right angles, so
a moving scanning beam can scan tags moving fast, slow, or not at all?
There must be a way to do this at a reasonable price!
Dave Gibbons
Brian74777
I recently saw a barcode system installed on a layout during layout tours
with the South Suburban Division of the NMRA (south Denver). While the
software was not yet controlling operations, the system could read the
code on every car as they entered or left a yard. The engines were not
coded. Other than the need for a non-prototypical 8 foot barcode label
(HO scale) on the bottom of the car, the system looked great.
Brian M.
email: brian...@aol.com
"If there were only one truth, it would be impossible to paint a hundred canvases
on a single theme"--Picasso
Daniel A. Mickey
The motion of the cashier at the market probably is not "regular" but it
works!
Dan Mickey
Rick Fortin
Another idea is to use the small tags that are inserted under the skin
of pets for electronic identification. These are the size of a grain of
rice, and are activated whenever the scanner is placed within a few
inches away. Each tag has a unique ID number. The small size of the
device can easily be hidden under a car or loco, even in 'Z' scale.
Unfortunately, the scanners are about $600 each (about like a bar code
scanner) but it may be possible to locate surplus scanners. The tags are
only a few cents each. The system is ultra reliable and does not suffer
some of the disadvantages of bar coding (think about how many times it
takes to get a reliable scan on some items at the grocery store).
Still another technology which may possible to convert is the proximity
readers often uesed to allow employees into secured buildings or areas.
This is similar to the pet tag devices, but are a little larger, work
differently and are a bit cheaper. These are easier to find at surplus
electronics places.
Rick
Rick
Steven Dorst
On Fri, 02 May 1997 17:23:33 GMT, I wrote:
SNIP
>Size of barcode: .....
SNIP
>envision using 6. 5 and 1 check digit. This works out to 4 digit
OOPs, make the last of the above line ...3 digit
>lenghts (remember the Interleave). The five digit serial number
SNIP
Sorry, My original calculations were based on a 7 digit ID with the
owner ID being 3 decimal digits. I recalculated when I realized that
the owner number didn't have to be 3 decimal digits, 8 binary digits
would do just fine, and I didn't correct the resulting digit length in
my original summary posting.
Steven Dorst
On Thu, 01 May 1997 18:36:37 GMT, I wrote:
>Hello,
>Our club is beginning to investigate the use of bar code tracking of
>rolling stock on our layout. We have 1800+ cars and 700+ engines that
>belong to members currently certified to run on our railroad. During
>our 4 hour public operating sessions, we run with 7 mainline cabs, and
>have 30-50 trains seen during a session.
>
>We envision putting a bar code on the bottom of every piece of rolling
>stock, then putting multiple readers at various critical points on our
>layout.
>
>I'd like to know if anyone else out in r.m.r land has experience in
>using bar code technology in conjuction with model railroading.
>
>Thanks in advance,
Wow, 18 hours, 7 responses (2 by email), plus some web-browsing by
myself. Here's a summary:
1 person with experience barcoding in a non-model railroading
environment offered help and his phone number. I have NOT included his
name here as he emailed me and didn't post his reply to the net. Note
to MH: This is a long term project for our club and we are far from
the point of needing programming assistance, but I'll keep your number
on file for a year or three when we are ready to proceed.
Fritz Milhaupt <milh...@lma.rust.net> wrote that he envisioned using
the barcodes on waybills that go with the cars, not on the cars
themselves so that data would be entered OFFLINE after the run
session.
"Daniel A. Mickey" <dmic...@ix.netcom.com> wrote with excitement
about giving the software engineers something new to work on in the
hobby. He also wondered about the costs per reader.
jpur...@nyx.cs.du.edu (John Purbrick) wrote (in response to Fritz
Milhaupt) that bar coding the cars themselves gives much greater
flexibility in how they are used. John also worries about price,
noting that the bare optics can be obtained for about $25, but that
this would then require designing the supporting logic and data
transmission. He also questions the reliability of the car motion for
reading a bar code.
"Daniel A. Mickey" <dmic...@ix.netcom.com> responded to John that the
motions a cashier makes to read a bar code seems to be wobbly but just
fine. <I'll have more on this later>
brian...@aol.com (Brian74777) writes that he has seen a layout in
South Denver with a system in place at the end of their yards. No
software support yet, but the system does read the codes from the
cars. Brian (or anybody else): Tracking down the owner(s) of this
layout would be invaluable!!!
The second email response (JO) offers help, and concern regarding
length of bar code and whether the railroad could run slow enough for
a reliable read. Note to JO: like MH above, I'm filing your offer of
help in a place where I know I'll get to it at the appropriate time.
Results of my web browsing:
First, I didn't spend much time on the web, but I did find one company
with lots of background information on line, the web address is:
http://www.csensors.com/help_page.html
Most of the following information is based on their information.
Cost: They don't list prices, but at least one of the above responses
mentioned CCD readers at $150 each. While I paled at the price, there
is another reason I won't use CCD wand style readers and that is that
they require CONTACT with the bar code. I would never put a fixed
device between the rails that would protrude above the railhead, it's
an invitation to disaster. Also, the mounting height of the barcode
will vary significantly from VERY close mounting on the fuel tank of
diesel locomotives, to mounted on a styrene flat piece mounted between
the hoppers of a gondola. Since these are differing heights, CCD's
seem to be out of the picture.
Technologies: There seem to be 3 technologies of bar code readers.
CCD's drawbacks are mentioned above, the second is the laser reader
commonly seen at supermarket checkstands. These use a SINGLE laser
that is rapidly moving back and forth to produce a sufficiently FAST
speed of scanning, causing us to perceive the laser as a line.
Supermarket checkstands that seem have multiple lines are constructed
by dividing the beam of a single laser and using mirrors to give them
different orientations so they can read codes at any angle. The back
and forth motion of the laser would be difficult to accomodate between
the tracks, at least in HO Standard guage. Another drawback to lasers
is that they are visible if there is enough particulate matter in the
air above them. As our club does night operations, I am forced to rule
out standard lasers as a reader. The third technology is the fixed
beam, LED lighted scanner. They can reliably read from 3/8 inches to
3/4 inches away from their source. They (like lasers) are available in
the high-infrared region that is invisible to the naked eye. They
(unlike lasers) have a much lower risk of eye damage if you accidently
look directly into them.
Speed of scanning: Based on the specs of the one fixed beam led model
I found, they can read a code at 6-60 inches per second, this
translates to between 30 and 300 scale miles/per hour in HO scale,
obviously much to fast for use anywhere but on a high speed main. I
intend to write this company as well as others to see if slower speeds
can be managed.
Size of barcode: First, barcodes do not depend on the height of the
bars, only on the width of the bars and spaces between the bars. Tall
bars seen on packaging are big to allow inaccuracies in aiming a
(handheld) laser scanner at them from some distance away. The height
of a barcode can be made extremely short IF you can assure that the
reader will always be correctly positioned, which I don't think will
be a problem with centerline mounted barcodes fixed to cars in a fixed
guideway (the track). Second, the length of the barcode. Of the
various barcodes, the one with the greatest information density is the
Interleaved 2 of 5, which is a numeric only code that uses the bars to
code one digit, and the spaces to code the next digit. Interleaved 2
of 5 code always contains an even number of digits, of which I
envision using 6. 5 and 1 check digit. This works out to 4 digit
lenghts (remember the Interleave). The five digit serial number
(00000-99999) would encode 256 owners by 390 cars/owner. (The few
people with more than 390 cars would be given a second owner ID) I'll
leave the math out, but this results in a length of approximately 0.64
inches including stop and start digits, or just over 4 1/2 scale feet
in HO, which should be mountable on any car.
Communications:
Bar code readers seem to come with three possible communications
options: keyboard substitute, RS232 serial, and RS422/485 serial. The
serial options can also incorporate polling where the controlling
computer asks for data rather than the reader blindly sending the
data. Keyboard substitute would be fine if we wanted just 1 reader,
but that would defeat most of the purposes of barcoding. RS232 is
better, but multiple readers require multiple serial ports. RS422/485
readers can share a single computer port if you use the polling method
of data collection.
Summary: Based on VERY preliminary research, I'm leaning toward a
fixed beam, infrared LED device using polled RS422/485 communication
between multiple readers and a single port on a single computer.
Interleaved 2 of 5 code will be used for it's compact length. All of
this assumes that readers can be found at a reasonable price (or
possible donated, our club is a non-profit, educational institution).
I haven't started thinking seriously about software yet, but switch
list generation, car and train tracking, production of a prototypical
train profile sheet and manifest, as well as bad-order tracking seems
to be just the beginning.
As more responses come in, and as I do more research, I'll keep the
group posted.
Lou (Soo) Papineau
"Daniel A. Mickey" <dmic...@ix.netcom.com> wrote:
->How much are bar-code readers?
In the Newark catalog #114 page 196, they list the wands and the
PC interface, but don't list prices.
For a catalog call: (312) 784-5100 or visit
http://www.newark.com/REQSTCAT.HTM
Lou (Soo) Papineau
New Haven Society of Model Engineers
Celebrating 65 Years of Model Railraoding
DLChambers
>> Should tags run parallel to the length of the car. or a right angles,
so
>> a moving scanning beam can scan tags moving fast, slow, or not at all?
Given that the readers shoot the beam along the length of the barcode, I'd
suspect that it would work better if the barcodes ran parallel witht the
rails (meaning, that the individual bars of the code run perpendicular to
the rails)
I've been interested in this myself for a while, but have drawn similarly
disappointing conclusions about the necessity of using a laser reader, and
it's price.
I have considered this: Use a "wand" reader (which are relatively
inexpensive), position it to be at rail height, place the barcodes on the
bottom of the trucks, hovering 1/8" above railhead. Place the same barcode
on both trucks of each car to give thr reader a "second chance". You'd
need to test if the wand-type reader could reliably read at 1/8" distance;
it's intended for reading when the tip is in contact with the printed
barcode. I suspect that in this configuration train speed wouldn't be an
issue; wand readers can read faster than the train can move.
-Dave
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
Opinions expressed are my own, unless you agree, then they're yours, too!
I'm a member of the Palo Alto Model Railroad Club http://www.pamrrc.org
RPearce900
In article <19970502175...@ladder01.news.aol.com>,
dlcha...@aol.com (DLChambers) writes:
>Given that the readers shoot the beam along the length of the barcode,
I'd
>suspect that it would work better if the barcodes ran parallel witht the
>rails (meaning, that the individual bars of the code run perpendicular to
>the rails)
>
>I've been interested in this myself for a while, but have drawn similarly
>disappointing conclusions about the necessity of using a laser reader,
and
>it's price.
>
>
Hi guys,
I have been considering this matter for a while myself. I have some rather
optimistic plans for my layout which would make some automation almost
essential, and that would NEED to have some way of identifying precisely
what was passing over a particular point on the track.
Ok, so we pretty much have to reject commercial barcode readers with their
laser technology, on the grounds of cost. But don't give up, I believe
what we need can be done simpler.
For example, I don't think we really need a 10 (decimal) digit code on
each wagon. My guess is that 8 bit (0-255) would be enough for most
purposes. By placing the barcode along the direction of travel, each bit
can be a quarter inch long and still fit most HO stock. At that scale, a
reflective photodiode detector could be used.
Now to make the reading method easier, we add a second detector and print
a 'clock' pulse train beside the real thing. By offsetting the blobs it
becomes possible to detect not only direction of travel but also
orientation. Sketch diagram time (use monospaced font) .....
== == == == == == == == ==
== == == ==
Hmmmm, did that make any sense ?
Oh well, just some thoughts.
Rob
Stanley Conley
Hello all;
I was cruising a consigment/junk shop yesterday and lo and behold! a barcode
operated robot train set. the track was two rail plastic , but elevated for
the most part, the train was a three unit affair modeled after a park people
mover type of setup. There was a single loop of track with a station on it.
Supplied with the track were several dominos to be placed bewteen the rails,
there were barcodes on the top of the dominos and provided instructions to the
train such as stop for 5 seconds, stop, wait and reverse, loop 5 time and stop
for 5 seconds, etc. The box graphics looked kind of seventiesish, This was up
for sale for about $75, I manage to restrain myself and still have my $75!
So if it can be made to work as a toy, should be more than possible at a
reasonable cost to work on a layoout.
stan
Stanley Conley Department of Civil and Environmental Engineering
Email: sco...@ccs.carleton.ca Carleton University, Ottawa, Canada
Mick Racer
>> Should tags run parallel to the length of the car. or a right angles,
so
>> a moving scanning beam can scan tags moving fast, slow, or not at all?
Will the reader beep every time a car is scanned? As for misreading,
instead of putting duo ID tags on each car, the computer driven software
could back the train up and re-read the defective car. If, after three
times, the laser is unable to read the car, then an adjacent bad-order
track next to the laser would be for unreadable cars. This would help
allieviate the problem of crashing the software.
Robert Heller
rpear...@aol.com (RPearce900),
In a message on 4 May 1997 22:35:17 GMT, wrote :
R> In article <19970502175...@ladder01.news.aol.com>,
R> dlcha...@aol.com (DLChambers) writes:
R>
R> >Given that the readers shoot the beam along the length of the barcode,
R> I'd
R> >suspect that it would work better if the barcodes ran parallel witht the
R> >rails (meaning, that the individual bars of the code run perpendicular to
R> >the rails)
R> >
R> >I've been interested in this myself for a while, but have drawn similarly
R> >disappointing conclusions about the necessity of using a laser reader,
R> and
R> >it's price.
R> >
R> >
R>
R> Hi guys,
R>
R> I have been considering this matter for a while myself. I have some rather
R> optimistic plans for my layout which would make some automation almost
R> essential, and that would NEED to have some way of identifying precisely
R> what was passing over a particular point on the track.
R>
R> Ok, so we pretty much have to reject commercial barcode readers with their
R> laser technology, on the grounds of cost. But don't give up, I believe
R> what we need can be done simpler.
Digi-Key sells some reflective photosensors:
(from Page 447 of their April-June 1997 catalog)
QT Optoelectronics Reflective Photosensors
Dia- Circuit Output Min/Max Optimim Termination Digi-Key Pricing
gram Type Output Sensing Part No. 1 10
Ic@If Distance
S - Transistor .6/3@40mA .150 Leads QRB1114QT-ND 1.58 14.81
S - Transistor .6/3@40mA .150 Wires QRB1134QT-ND 2.83 26.56
T T1 Transistor 1/5@20Ma .050 Leads QRD1114QT-ND 1.50 14.10
Diagram S is a wedge shape, much like a stealth fighter, with four wires or leads
coming out the back. Diagram T is a like a little DIP chip with 4 leads and the
sensor and LED side by side.
They also have a similar set of devices from Omron, much the same as the QT
device, but slightly more expensive.
A device like this and a self-clocking bar code should be fairly cheap. The
self clocking does not need two side-by-side bar codes. A single set of bars
is enough. The trick is variable *width* bars. This can encode both a 1/0
and an implied clock pulse. A simple method is like this:
Here is a '1' bit:
|---------------------------------| 'H' (== black)
| |
-----------------| | 'L' (== white)
| 1/3 bit time | 1/3 bit time | 1/3 bit time |
"bit time" == bit width == 1 bit's worth of car length
Here is a '0' bit:
|---------------| 'H' (== black)
| |
-----------------|-----------------| | 'L' (== white)
| 1/3 bit time | 1/3 bit time | 1/3 bit time |
Note: each bit takes the same length, but '1' bits are fat and '0' bits are
skinny. The serial to parallel logic would just be a clock separator (senses
black to white transitions), a 16x PLL, and a UART. The clock separator and 16x
PLL generate a variable 16x clock for the UART. The UART takes 16 samples and
counts H samples, if 8 or more, it shifts a 1 into the output, else a 0. You
would need a special set of guard bits at the start and end of the bar code to
get synchronization and to detect bit ordering (i.e. direction of travel).
R>
R> For example, I don't think we really need a 10 (decimal) digit code on
R> each wagon. My guess is that 8 bit (0-255) would be enough for most
R> purposes. By placing the barcode along the direction of travel, each bit
R> can be a quarter inch long and still fit most HO stock. At that scale, a
R> reflective photodiode detector could be used.
Right.
R>
R> Now to make the reading method easier, we add a second detector and print
R> a 'clock' pulse train beside the real thing. By offsetting the blobs it
R> becomes possible to detect not only direction of travel but also
R> orientation. Sketch diagram time (use monospaced font) .....
R>
R> == == == == == == == == ==
R> == == == ==
R>
Don't actually need this (see above).
R> Hmmmm, did that make any sense ?
R>
R> Oh well, just some thoughts.
R>
R> Rob
R>
--
\/
Robert Heller ||InterNet: Hel...@CS.UMass.EDU
http://vis-www.cs.umass.edu/~heller ||FidoNet: 1:321/153
http://netmar.com/mall/shops/heller /\
Brian74777
>
> What was the max speed of the cars at which the barcode reader would
> continue
> to read reliably? Any idea of the cost?
> F. Alpaiwalla
> email: alpai...@indy.tce.com
>
I can't address the speed issue, though they were installed at the yard
throat entrances so I assume the trains wouldn't bee going very fast. He
said the scanners were just standard commercial scanners so I assume they
wouldn't be too bad--It seems to me that the backs of the computer
magazines have ads for such things. He had it all hooked up to a PC and
was writing the application in what he said was a stardard database
program. He also said the PC was not particularly high performance. That
end shouldn't be too bad.
DLChambers
rpear...@aol.com (RPearce900) wrote:
>> My guess is that 8 bit (0-255) would be enough for most purposes.
Not even close. We're a club, and one of our members has > 300
locomotives!
>> Now to make the reading method easier, we add a second detector and
>> print a 'clock' pulse train beside the real thing.
Interesting idea. Data density would be an issue, but it could work and
would definitely be inexpensive!
Andy Robinson
On Thu, 01 May 1997 18:36:37 GMT, do...@usa.net (Steven Dorst) wrote:
>
>Hello,
>Our club is beginning to investigate the use of bar code tracking of
>rolling stock on our layout. We have 1800+ cars and 700+ engines that
>belong to members currently certified to run on our railroad. During
>our 4 hour public operating sessions, we run with 7 mainline cabs, and
>have 30-50 trains seen during a session.
>
>We envision putting a bar code on the bottom of every piece of rolling
>stock, then putting multiple readers at various critical points on our
>layout.
>
>I'd like to know if anyone else out in r.m.r land has experience in
>using bar code technology in conjuction with model railroading.
>
>Thanks in advance,
>
>
>Steven J Dorst <do...@usa.net>
>HO Scale Trainmaster
>East Bay Model Engineers Society at the
>Golden State Model Railroad Museum
>Visit us on the web at:
>http://www.hooked.net/users/robertjm/gsmrm.html
For those who are interested:
Psion series 3a barcode reader manufactured by Paradigm called the 3
wand and comes with software for writing the code to a file. Price in
RS components catalogue is £299 ($475)
CCD Bar code readers (panasonic) (keyboard input) are £260 ($400) +
£39 ($60) for the PC interface.
Barcode wands by HP (5V serial asynchronous interface) are £229 ($360)
Andy Robinson
an...@digging-deep.demon.co.uk
http://www.digging-deep.demon.co.uk/railway/
DLChambers
an...@digging-deep.demon.co.uk (Andy Robinson) wrote:
>> Psion series 3a barcode reader called the 3 wand...
>> Barcode wands by HP...
Andy,
Wand-type readers will probably not work. They rely on the tip of the wand
to be pressed against the barcode as the wand is moved across it. The
optics in the tip of the wand are designed to "focus" on the tip. If, for
example, you drilled a hole thru the roadbed between the rails and mounted
the wand pointing upwards (toward the bottoms of the cars), I doubt it
would work. The 1/4" to 1/2 between the wand tip and the barcodes on the
bottoms of the passing cars is probably too great a distance for the wand
to read the barcodes.
I have a wand-type barcode reader. This weekend I'm planning to set up the
experiment described above and post my results here on Monday.
John Purbrick
In article <336f13fd...@news.demon.co.uk>,
Andy Robinson <an...@digging-deep.demon.co.uk> wrote:
>On Thu, 01 May 1997 18:36:37 GMT, do...@usa.net (Steven Dorst) wrote:
>
>>
>>Hello,
>>Our club is beginning to investigate the use of bar code tracking of
>>rolling stock on our layout. We have 1800+ cars and 700+ engines that
>>belong to members currently certified to run on our railroad. During
>>our 4 hour public operating sessions, we run with 7 mainline cabs, and
>>have 30-50 trains seen during a session.
I think barcode wands are an unworkable idea. They require contact with the
object being scanned and cost hundreds of dollars. A scanning-laser reader
as used in supermarkets would maybe work but the cost is still prohibitive.
Much better would be an integrated focused receiver-transmitter component.
These have internal lenses so that a light beam is sent out and received,
bouncing off a small area of a surface a certain distance from the face of
the component. If you can get hold of the Hewlett-Packard Optoelectronics
data book, you will find several listed. The least expensive one is the
HBCS-1100, which has a standoff distance of .168 inches and a modulation
transfer function of 70% at a line spacing of 2 line pairs per mm. (MTF
is the proportion by which the maximum output is reduced when the input
goes from full white to full black. A line pair means one white and one
black line.) But note that if there is a .020 inch error in depth of the
target, output falls 50%! Newark Electronics lists the HBCS-1100 for $27.00.
Another manufacturer is Opto Technology: their component is the OTR680, but
it has a large structure whereas the H-P version fits in a round case just
over 1/4 inch in diameter. They aren't nearly as informative in their catalog
either and I don't know the price.
John Purbrick
Lou (Soo) Papineau
Check out:
Lou (Soo) Papineau
New Haven Society of Model Engineers
Celebrating 65 Years of Model Railroading
DLChambers
sco...@ccs.carleton.ca (Stanley Conley) wrote:
<<<
...
a barcode operated robot train set.
...
there were barcodes on the top of the dominos and provided instructions to
the
train such as stop for 5 seconds, stop, wait and reverse, loop 5 time and
stop
for 5 seconds, etc.
...
So if it can be made to work as a toy, should be more than possible at a
reasonable cost to work on a layoout.
>>>
Actually, that's a bit simplistic. Sounds like those "barcodes" are
extremely simple, maybe a total of 10 or so "commands", so the set of
patterns that the reader needs to be able to discern/interpret is small
and therefore very simple.
That's very different from reading 16 bits from a barcode off a car moving
at speed, and doing it for each car that passes by.
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
Opinions expressed are my own, unless you agree, then they're yours, too!
Member of the Palo Alto Model Railroad Club since 1974
Take a tour of our Club at http://www.pamrrc.org
Grizzgear
In 1978, my brother made a scanner out of an LED and a light sensative
diode. We put modified barcodes on the bottoms of a few cars, and were
able to read it. The barcode plates were binomial, with alternating black
& white stripes down the center (to tell the reader to look for new data).
On either side of the center stripe were binomial code strips (two of
them make the car reversible), white=1, black=0.
We were successful in reading cars. The binomial card on the bottom of
each car was about 0.5" x 1", which was cumbersome for [N] scale. And,
the distance from the under-track reader had to be fairly constant, so the
undercar details were sometimes problematic (hopper bays got chopped).
Problem was, we couldn't afford a computer to do the yard processing at
that time. Now you could do it all with an old 286!
Spencer Craig, Pres., Grizzly Gear
Your Source for Innovation in Camping!
http://members.aol.com/grizzgear
Joe Ellis
OK... here's some data from a retail perspective.
I work at a large sporting goods store. The registers use scan guns to read
UPC bar codes. I did some experimenting during a slow period today.
The gun will read a standard UPC bar code at distances from contact to 10
inches. The object can be moving at at least HO scale speeds (judged by
passing the product thru the scan beam). Since it reads in a straight line,
at right angles to the bars of the code, even a center-beam application on
a N-scale car is feasible.
The gun we use is a Symbol Technologies LS 2000. I do NOT know how much
they cost... I'm not in that end of the business... but I do know they've
been around a while. We've had ours for over 3 years.
With a under-rail mount lengthwise between the rails, this should be usable
in nearly any scale for main yard applications. Whether it's useful for
sidings would depend on the cost.
--
Delete "NOSPAM" in address above before replying!(fraggin' morons...)
Joe Ellis o/~ The Synthetic Filker o/~ | TesserAct Studios
\\ __\\\__ ell...@mail.idt.net | 1800 W. Kemper Rd
()XDarwin(;> Now on the Web at | Cincinnati, OH 45240
/ /~LL~~LL~ http://shell.idt.net/~ellis69 |New Dimensions In Filk!
Unsolicited commercial E-mail will be proofread at $25 hr/2 hr min.