Remote data collection using the Internet

Talk to be given at the American Association of Physics teachers meeting in Anaheim CA, Jan. 9-14, 1999.

Who we are:

Ray Wisman (Computer Science) and Kyle Forinash (Physics) at Indiana University Southeast.

With IUS students: Lori Blankenship (who actually did most of the work), Mike Riley, James Phares

What we are up to:

Problem: Suppose you wanted to measure a local quantity (temperature, pressure, radiation level, magnetic field, sunshine intensity, etc.) at many different locations across the world, all simultaneously in real time?
Solution(s):
- A) Write a research grant for big money; spend accordingly, OR
- B) Find accomplices with computers on the internet at the sites, attach appropriate probes to the computers and access the probes remotely (basically for free).
Method: Solution B! We asked people on the Phys-L list serve of physics teachers to go to our Registration Page if they were interested. More than a dozen people responded (although only three sites actually registered to participate).
Sites and times: Three groups off campus (Ticonderoga, NY, Omaha, NE and Edinboro, PA) volunteered to participate over the Christmas break. We added four computers on the IUS campus in New Albany, IN.
Hardware: Each remote computer had a ULI interface box from Vernier (connected to the serial port) with a light probe attached to it. We chose the light probe as a simple proof of concept device (any probe would have worked).
Software: Each remote computer ran a small utility program called TCP2Serial which allows a direct connection between the TCP/IP internet connection and the serial port. The data collecting computer (here on campus but which could also have been anywhere on the internet) ran a program that could talk over TCP to tthe remote computers serial port, sending comands and retrieving data from the ULI. Two students did the programing as a class project: Lori Blankenship wrote a stand alone JAVA program which reaches out over the Internet and controls multiple ULIs simultaneously (program inputs: IP numbers, sampling parameters; output to local hard disk: data from the ULI). James Phares worte a Visual Basic version of th eTCP to Serial communications program.

What kind of results did we get?

To see some of the actual data we collected with comments go to the second results page.

Project Summary:

Easy to do (software was written by students).
Software is generic (doesn't care which analog probe is attached).
Concept will work with ANY serial device (not just the ULI) and ANY probes (temperature, magnetic field, radiation, pressure, etc.).
- Basic Stamp
- AdcDongle
- ULI
- MPLI
- CBL (serial link under development)
Inherently democratic (anyone can access the network of probes at any time for any purpose).
Inexpensive (around $300 for the ULI and probe).
Potentially very powerful (for example: it would be easily be possible to have a finer temperature monitoring grid across the US than can currently be achieved by the National Weather Bureau).

What could this be good for?

Suppose a large number of computers on the internet had several probes hooked up (for example, radiation, light intensity, magnetic field, etc.) and were left on all the time. An internet user in California might decide to sample the noon solar intensity every day in one county in CA for an entire year by taking data from the computers in that county which are participating. Another user in Italy might study the variation of solar intensity across the US in a 24 hr period. Local and global variations in the earths magnetic field (if any) might be examined over short or long time periods by someone in Canada. Radiation spread from a nuclear test could be monitored world wide and correlated with weather changes. The neat thing is ANYONE who had a list of the locations (IP numbers) of the participating sites could do whatever study they wanted to without bothering the owners of the computers at all, as long as everyone agreed to have the same probes available.

Return to the IUS Physics Top Page.