The 2018 IoT Challenge by T-Systems raises the question of how environmental damage caused by flooding can be limited using the Internet of Things and Data Analytics. The winner was software developer Bin Zhao. His solution captures and analyzes the data in real time using CONTACT Elements for IoT.
Bin Zhao relies on a mix of open source software, standards and an open, modular IoT platform for his "Water Monitoring and Flow Measurement System ". His solution is based on a network of measuring stations that are simply mounted on bridges or masts. The sensor data is transmitted to CONTACT Elements for IoT in real time via the narrow band IoT module BC 95. The prognosis algorithm can be adapted remotely for each terminal device.
Zhao uses a proven standard from the AI for prediction: the Long Short-Term Memory Model (LSTM), which supports machine learning in multi-layered neural networks. In addition to real-time data, Zhao uses average values from the past months such as Rhine water levels, rainfall and snow melt to train his model and obtain reliable flood forecasts.
Reliable monitoring of Rhine tributaries helps the Federal Government and the states to identify imminent flooding and to initiate measures for public safety and environmental protection at an early stage. Such data are also relevant for weather services, insurance, agriculture or smart metering, for example.
Bin Zhao (27) is a developer for IoT applications at CONTACT Software.