The LoRaWAN pilot will make Launceston the first Australian city to have a city wide IoT network.
Additionally the LoRaTAS team has made it an open innovation data infrastructure for Launceston so that everyone from schools to communities, small businesses and government can participate in by using IoT enabling technologies to develop applications to solve city problems.
Some examples include:
- real-time traffic management (like tracking buses, taxis etc),
- real-time environmental monitoring (monitoring air-quality, water levels and fire management),
- state of the art sensor management on farms and vineyards,
- innovative solutions for solving problems in health (fall management and dementia tracking), and
- even simple interactive systems like binary voting for citizens.
Open City Data
Smart learning centres
The hierarchy of application themes for the pilot
Sensing of some sort is the foundation to all our pilot applications. However, some applications only need to collect values that a human will interpret and understand without any other "smarts", and so this the most basic project type in the pilot.
Predictive applications extend beyond a basic sensing application by using algorithms or machine learning to predict something from the measurements. A common application is to this for predicting failure in equipment so that appropriate maintenance can be scheduled in time to avoid failures. Prediction can be generalise to a wide range of applications involving risk, from electrical equipment maintenance, to crop disease risks. While prediction applications are most commonly determine a level of a risk it is still a human that is informed of the prediction and the human acts on it. If the action (response to reduce the risk) is also by a machine then it is a Control Application, which we will talk about next.
A control application is one which uses both sensing and the prediction to formulate an action that is carried out automatically. Another way of thinking about it is that control applications send out commands that turn remote devices on or off, or speed them up or slow them down while predictive applications only collect the sensing data to process without controlling the remote devices.
The peak of the application hierarchy extends beyond the automatic Control Application by including input from humans to guide the actions that the system would have otherwise automatically taken. As an example, imagine a home that had an automatic energy management system to heat the house only when needed, and it detects the person in the room and knows their preferred temperature so it automatically sets that temperature for them, then imagine the person is exercising and for a short time wants the temperature to be lower, but not to change their settings. It is just a little "dip" to the temperature for a short period so they don't overheat and can exercise in comfort. That is a simple example of human interaction modifying the response that would have been followed automatically by the system by including a factor the human understands that is beyond a "permanent" setting for the system.