ul 2004
—M MM
vector
visible
into an
| io the
e area.
g Java
mation
licating
w form
Tent or
Ir users
This is
service
pecific
ication
". Aeris
nboard
eration
as well
sor can
rations
anually
cations
arming
engine
figured
ed into
jue for
ecurity
cation
-based
y and
atform
:les for
seen in
could
| assets
Asset-
ughout
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV. Part B2. Istanbul 2004
North America. During the tests, pages were issued to the
Asset-Link sensors for their locations. Their locations were sent
back to the server and instantly updated into iVCAMS3 and
plotted onto the map. The accuracy of the location data was
confirmed with the drivers of the cars and was high. Other
commands, such as remote unlocking of doors, were also tested
and were successfully received by the sensors and the correct
operations performed.
The total time necessary to complete all steps in paging an asset
were also monitored during the course of tests. The times for 29
pages are shown in Figure 2. These pages were sent to two
assets; one asset traveling between Calgary and Edmonton
while the second asset was in Fort Lauderdale, Florida. The
average time for roundtrip of a page was found to be 9.25s,
with a maximum time of 16s, which was reached by only one of
the pages. The standard deviation was two seconds.
Roundtrip Latency of Pages to Assets
Latency (s)
S
+
+
+ * + ++ ++
g E iE [5 20 ΠE
Figure 2. Page latencies
4.2 System Testing
To test the theoretical limits on the system, in terms of number
of assets that can be supported, simulation testing was
performed using several different computer systems acting as
the Server. Simulation enabled precise control of variables and
made it possible to quantify the limits of the Platform. In its
development tools, Aeris provides software that simulates the
Aeris AS and DS servers. This enables developers to create
applications that are fully compatible with the Aeris protocol
before using it with the real Aeris system. The simulated server
is customizable and can be modified to send messages at any
defined rate. In this way, we can simulate a different number of
units by varying the frequency that messages are sent out by the
simulated Aeris server. We can then determine the limitations
of the system in terms of the number of units and assets that it
can support, by monitoring the CPU and network bandwidth
usage using PerfMon. These monitored attributes arc displayed
in Figure 3. The server tested consisted of an Intel Pentium III
933MHz system with 512MB of RAM and running the
Microsoft Windows 2000 operating system.
With an assumption that an asset will send a message once
every 30s, 64 messages per second approximates to a fleet of
about 1900 assets and 131 messages per second represents
about 3900 assets. From the results of Figure 3, the CPU usage
for even 3900 assets is not taxing the server, as it only ranges
from 20% to 25%. For 1900 assets, the CPU usage is lower as
would be expected, ranging from 7% to 15%. As for network
bandwidth usage, the simulation results shows that this should
not be a major concern as the usage is 15KB/s and 29KB/s for
1900 and 3900 assets respectively and well within the limits of
even home broadband Internet connections.
Simulated Asset Testing
FH oh ft fh be hd rh ng Adoni ht tr mith
CPU Usage (%)
KB/s Received
o c à
Time
Figure 3. Simulated asset testing
5. CONCLUSION
A fully complete platform has been developed and is capable of
addressing all major concerns in MAMS, including asset data
acquisition, transmission and handling and analysis as well as
unique capabilities for user access to asset data. It reduces the
resource, time and cost needed to deploy a MAMS, thereby
enhancing its benefits. It features a modular design separating
the important functions of a MAMS and enables developers the
flexibility to choose whether to use all or only certain
capabilities of the Platform.
This Platform was used as a basis for a real-world MAMS
known as iVCAMS3. This system was built based on the
requirements of industrial partners and field testing was
conducted to ensure that the asset could send messages and
locations back to the Office and that users could issue
commands to the asset. Testing was completed successfully
with vehicles with Asset-Links installed in multiple cities in
Canada and the USA. It was possible to monitor vehicles when
they were stationary and in motion. Roundtrip latency of
messages was around 10s, making it possible to have real-time
monitoring and control capabilities with assets.
ACKNOWLEDGMENTS
The research was supported by a NCE GEOIDE grant.
REFERENCES
Aeris.net, 2003. “MicroBurst Technical Details”,
http://www.aeris.net/aeris web/products microburst technical.
html (accessed 1 Feb. 2004)
CSI Wireless, 2003. “Asset Management Products”,
http://www.csi-
wireless.com/products/documents/New AssetLink.pdf (accessed
| Feb. 2004)
FCC, 2003. FCC: Enhanced 911.
http://www.fcc.gov/91 l/enhanced/ (accessed 14 Oct. 2003).
Lee, S. and Y. Gao. ^Mobile Asset Tracking and Management
Over the Web". Proccedings of Geodesy for Geotechnical and
Structural Engineering. May 21-24, 2002, Berlin, Germanv.
Peng, Z. and M. Tsou. Internet GIS — Distributed Geographic
Information Services for the Internet and Wireless Networks.
John Wiley & Sons, New Jersey, 2003.
Based Services",
Prasad, M., 2001. "Location
http://gisdevelopment.net/technology/lbs/techlbs003pt.htm
(accessed | Feb. 2004)
