Fostering of IoT Solution Towards Smart Water Management:
Case Urban Flood Prediction 

Developing IoT solutions requires high expertise and testing. Savonia’s role in AMBITIOUS project supports the development of an IoT infrastructure for AI-enabled prediction and data analysis services in smart water management. It aims to predict urban floods based on local precipitation and water level of stormwater wells.  

Savonia experts will collaborate and support Funlus Oy to implement this pilot project. Savonia’s expertise includes up-to-date knowledge on smart water management solutions based on the integration of real-time data from various sources to provide the desired services.  

In today's world, the frequency and intensity of natural disasters faced by cities are steadily increasing due to the impact of climate change. Especially, flood disasters occurring in urban areas can cause serious damage to infrastructure and social life. As a pilot implementation of the AMBITIOUS project, a flood monitoring and prediction system is being established to meet the needs of the city of Kuopio.  


Structure of the System 
The pilot is being built upon an AI model that predicts the probability of urban flooding, using data from sensors that measure water levels in stormwater wells in real-time, along with local precipitation levels. Finnish Meteorological Institute (FMI) weather data will be used along with sensor data for flood prediction. The prediction is based on current measurements and weather forecasts. 

We utilized the Elsys ELT Ultrasonic Industrial Distance Sensor, which operates on LoRaWAN technology [1]. Fifteen sensors will be placed strategically in specific locations within the city of Kuopio, where the risk of flooding is considered high. Images depicting the installation of 7 sensors are presented in Figure 1. The sensors were mounted on the covers of stormwater wells. In addition to ultrasonic distance measurement, the ELT sensors are equipped with temperature, humidity, accelerometer, and atmospheric pressure sensors, allowing for comprehensive monitoring and analysis of the environment. 

Figure 1. Installation of Sensors. 

In the implementation, data is being sent from sensors that measure water levels in stormwater wells in real-time. This part is handled by the stakeholder company, Funlus Oy [2]. Different Communication technologies are being utilized for the transmission of data. Initially, data is received from sensors using the LoRaWAN protocol [3]. Funlus shares this data with Savonia using the MQTT (Message Queuing Telemetry Transport) protocol [4]. This data is being stored on Savonia’s private server, ready for monitoring and processing. The data transfer is carried out in real-time using LoRaWAN IoT. Savonia will experiment with different AI prediction models and test them to validate the results. The communication structure is illustrated in Figure 2.

Figure 2. Communication Structure. 

Savonia is using the Thingsboard platform for real-time data monitoring. Thingsboard provides a comprehensive and user-friendly interface for tracking sensor data [5]. Figure 3 displays the data flow from the sensors on the Thingsboard monitoring screen, showcasing how the platform facilitates effective and efficient observation of environmental conditions.

Figure 3. Thingsboard monitoring screen 

The ongoing pilot project demonstrates the feasibility of using IoT technology for urban flood prediction. The sensors are successfully monitoring the water levels in stormwater wells and transmitting data in real-time using LoRaWAN. The integration with the Thingsboard platform enables continuous monitoring and provides valuable insights into the environmental conditions. This approach has the potential to improve urban resilience against flooding and mitigate the adverse impacts of climate change on urban infrastructure and communities.

 

References 

[1] https://www.elsys.se/en/elt-ultrasonic/ 

[2] https://www.funlus.fi/ 

[3] https://lora-alliance.org/ 

[4] https://mqtt.org/ 

[5] https://thingsboard.io/