Overview

• Temperature Sensor

  This was used in our system because it is a relatively important aspect of an animal's condition and the data is easy to record and store. The device used is a straight temperature to voltage converter (TC1046). More information about its exact specifications can be found on Prodromos's report. There were no drivers to write for the temperatures sensor since all of the necessary conversions were done in hardware. The value passed to the system could simply be compressed and stored.
  
  

• Moisture Sensor

  We almost added this to the system, but eventually decided against it for cost reasons. We later discovered that it was just as well that we hadn't invested in this sensor as the three we were already using took up too many hardware resources to be able to fit on the chip.
  
  

• Accelerometer

  We decided to use this device but not quite for the purpose intended. Instead of measuring an animal's acceleration in any particular direction we would simply measure its level of activity. Because we now didn't need to know the direction the animal was accelerating towards, it meant that we did not need to an expensive 3-axis accelerometer. Instead we used a two-axis accelerometer and Martin connected it up in such a way that it would count the number of times the acceleration on either axis went above 0. This count gives a rough estimate of the animal's level of activity.
  
  

• Proximity Sensor

  The proximity data is probably the most important animal monitoring information to acquire. Ideally the system would record the exact position of every animal in the environment at every instant. The ZebraNet project does pretty well in achieving this by sampling the GPS coordinates of animals every three minutes. This is, however, impossible to achieve with the limited resources available on the chip. The best the PSoC can do is record proximity. This actually gives us the most essential part of location data. Knowing where animals are is not quite as important as knowing how close they are to each other. This tells us a lot more about their behaviour. The challenge in proximity sensing is to establish what type and how much prox information is required for effective analysis of animal proximity. These pages examine and discuss the hardware platform and algorithms used to create the proximity sensing module. There is also some data from experiments conducted to test the accuracy of the proximity data retrieved in various different environmental conditions.
  
  

• Heart Rate Monitor

  This was a very nice idea and it would probably be essential in the final system. Unfortunately due to financial constraints and technology availability we were un-able to add this device.