System Architecture and Hardware requirements
In the past, environmental sensors were special electronic devices custom-made to perform the function of taking photos and recording bird calls. However, with advances in wireless technology and significant price reductions for both 3G devices and connection rates it becomes possible to make use of camera enabled 3G mobile phones to perform the same task. Options considered by the project:
- Use 3G (phone) sensors to communicate back to the QUT data collection server.
- Have 3G only on some phones and provide secondary data communication channels to allow sensors to connect from non 3G to 3G phones.
- Set up a habitat server using a notebook or to communicate with the sensors via wifi.
Environmental monitoring and sensing
Smart environmental acoustic scanning - Sensors could be configured to automatically adapt to scan the environment at different times to detect sound. If sound is detected at a particular time, greater recording can be done around that period. If sound ceases at a particular time or if a change in bird behaviour occurs, other time slots could be analysed. If recording is continuous, this can be done offline. Sensors can be programmed to change the time at which they record to scan for interesting activity.
Use camera facility of phones to support taking photos of nesting Lewin’s Rail. The camera can be configured to take pictures/short video periodically to monitor the habitat of Lewin’s Rail, triggered by activity events. The video can be used in the future for tagging reference or further image processing. Low cost surveillance kits can be used with a laptop computer as a base station with connections to the sound activated cameras. With the wireless and remote monitoring technology used, we will be able to monitor the environment with more flexibility, causing less disturbance to the habitat.
Visualisation of acoustic signals
Our goal is a virtual environment for collecting, analysing, modelling and visualising the real environment. The environment will allow scientists to live in their data and to collaborate through it (a form of augmented reality). Sensors will monitor the environment and can be remotely controlled. Sensor data will be combined with remote sensing data to yield detailed information about the environment and how it is changing. Sophisticated analyses will detect trends; recognise fauna, weather and human disturbance; and correlate changes over time. The incorporation of predictive models will enable future scenarios to be analysed and visualised.