As Oli said - this / these may have been mentioned:

- Putting two detectors a known distance apart and timing the period betwee=
n
activation. Could be reflectance, capacitance, light/sound etc. Use three
and you get delta velocity as well. Use 4 and you can see if it's increasin=
g
its braking or increasing its acceleration (subject to measurement
accuracies - a massive vehicle is unlikely to have very high delta
acceleration.

- Single camera viewing arrival of leading edge with program timing delta T
between points in camera field. Using Canon's toolkit for A series camera
you could just about get a $100 camera to time it for you.

- Correlation of signals at two or more points works and gives you a
continuous record. Simplistically you look at a sensor signal which is
monitoring the vehicle (sound, light, capacitance,...) and correlate it wit=
h
a time shifted signal from an identical sensor a know distance away. Maximu=
m
correlation (hopefully) occurs when delta T in processor is same as delta T
of a point on the train. Once you get first lock you need hunt around only =
a
small range each time.
This method was used (and may well still be) to measure iron sand slurry
flow as iron sand was pumped with water through large pipes going out to se=
a
to a tanker loading iron sand. Sand and water are delivered to tanker and
water is discarded (not surprisingly).

NZ is very good at exporting raw material for other people to add value to
and sell back to us. As  a bonus our iron sand has a very high Titanium
content. Makes use in a normal smelter problematic - DO NO TURN FURNACE OFF
:-). Continuous casting method is used.


  Russell McMahon
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.