The study

High-resolution hydrological data from remote areas, in real time

Effective water resource management is increasingly vital due to climate changes and environmental challenges. A novel cloud-based system using Thinger.io and NB-IoT technology enables real-time monitoring of soil moisture and water levels, providing high-resolution data from remote areas: the NB-IoT Hydrological Monitoring System.

This breakthrough offers an efficient, scalable solution for addressing long-standing issues in hydrological monitoring. The solution has been tested in two regions in Spain, using dense sensor networks to measure both soil moisture and water levels. The main goal was to develop an efficient, low-cost monitoring solution for challenging environments, providing real-time data on hydrological processes. By integrating NB-IoT with the Thinger.io cloud platform, the system overcame traditional issues like low connectivity and high energy consumption.

The same network is on this site as a case study, UPM and UC3M: hydrology measured sensor by sensor, with the findings of the study in numbers.

The problem

What does it solve?

Traditional hydrological monitoring struggles with sparse data collection, low spatial resolution and connectivity issues in remote areas.

This new system, developed in collaboration with the University Polytechnic of Madrid and the University Carlos III of Madrid, along with Thinger.io cloud integration, provides a comprehensive and real-time solution to these challenges. It enables timely tracking of hydrological events, thus optimizing water management and aiding in flood prevention and agricultural planning.

Map of Spain and the Community of Madrid locating the field site, with an aerial view of the two catchments: a grid of red dots for the soil moisture sensors and a blue triangle for the water level gauge
Field site location and sensors layout: red dots are the soil moisture sensors, the blue triangle the water level gauge.

Why it matters

What real-time hydrological data is used for

Real-time hydrological monitoring has become essential for addressing challenges across various sectors.

Precise data on soil moisture and water levels improve water resource management, thus helping to optimize agricultural irrigation, prevent floods and create more sustainable urban plans. By constantly tracking these processes, authorities can implement effective strategies to mitigate climate change impacts, such as droughts and extreme weather events.

Climate science

Early warning systems for extreme weather events.

Water management

Efficient irrigation systems to reduce water waste.

Agricultural monitoring

Smart farming technologies to optimize irrigation.

Flood prevention

Infrastructure for flood prevention in vulnerable areas.

Urban planning

Sustainable water management for smart cities.

Drought mitigation

Efficient water use during droughts to prevent waste.

Key results

What the field trials showed

The system demonstrated excellent performance in field trials, with sensors providing reliable data transmission and optimized battery life through Thinger.io.

Key findings include:

  • High spatial and temporal data accuracy.
  • Successful operation in remote areas with low connectivity.
  • Extended battery life with efficient energy consumption management, optimized via Thinger.io.
  • Real-time data accessible via Thinger.io's cloud platform for immediate analysis.
Diagram: a soil moisture sensor network and an ultrasonic sensor network in a pipe both sending sensor measurements to a cloud platform for data storage and visualization
Scheme of the network's architecture.
Four panels: an ESP-32 board on a desk, a soil probe with its cable and breakout board, the ADC port schematic with a 10K resistor and capacitor on GPIO25, and the polar beam pattern of the ultrasonic sensor
ESP-32 chip (a), water level devices probe (b), ADC port schema (c) and beam pattern (d).
The two networks of the study, soil moisture and ultrasonic water level, report to the same cloud platform; the hardware behind each node, from the ESP-32 board to the beam pattern of the ultrasonic sensor.

The hardware

An NB-IoT node built on the Thinger.io toolkit

The core of the devices is an IoT development toolkit manufactured by Thinger.io, which gives the flexibility to create hardware prototypes tailored to the requirements of a project.

The toolkit is built around an ESP-32 System-on-Chip (SoC) processor, which combines Wi-Fi and Bluetooth capabilities, and also integrates a modem with NB-IoT capability from Quectel. The processor manages data acquisition via general-purpose ports with 12-bit SAR ADC functionality in order to collect data from the soil moisture and water level probes.

Catchment 1: a gently sloping field of low vegetation with young trees in tube guards, modern apartment blocks on the horizon under an overcast sky
Catchment 1
Catchment 2: rolling grassland with rows of saplings in pale green tube guards and a low hill in the distance
Catchment 2
A small black enclosure lying on sandy soil among weeds, with its cables running to the buried probes
Soil moisture sensor
Looking into an orange corrugated pipe from its mouth, with the ultrasonic sensor and its cable fixed to the top of the pipe wall
Water level sensor
The two catchments of the study and the sensors deployed in them: a soil moisture node on the ground and the water level sensor mounted inside the outlet pipe.

Two kinds of sensor

  • Soil moisture sensors: equipped with two independent probes wired through a Resistor-Capacitor (RC) circuit, allowing for reading calibration and analog noise reduction.
  • Water level sensors: use an A02YYUW ultrasonic sensor operating at 40 kHz, therefore providing sufficient accuracy to measure distances between 25 and 30 cm.

The units are powered by 3.7 V, 3400 mAh rechargeable lithium batteries. Solar panels were not used, to allow more flexibility in field deployment.

The node at a glance

Core
IoT development toolkit manufactured by Thinger.io
Processor
ESP-32 System-on-Chip, with Wi-Fi and Bluetooth
Connectivity
NB-IoT modem from Quectel
Data acquisition
General-purpose ports with 12-bit SAR ADC
Soil moisture
Two independent probes wired through an RC circuit
Water level
A02YYUW ultrasonic sensor, 40 kHz, 25 to 30 cm
Power
3.7 V, 3400 mAh rechargeable lithium battery
Solar panels
Not used, for more flexibility in field deployment

The Thinger.io NB-IoT toolkit board seen from above: USB-C connector, ESP-32 module, Quectel NB-IoT modem and SIM slot
Hardware architecture: the toolkit board, with the ESP-32 SoC and the Quectel NB-IoT modem.

The platform

How Thinger.io is used

Thinger.io played a crucial role in managing the data collection and transmission from the sensor network.

The platform is used to:

Sensor auto-provision

Each connected sensor is automatically provisioned, streamlining device deployment, data storage and device management.

Configurable sleep times

The platform enabled remote management of sensor sleep cycles to optimize battery life.

Data transmission and visualization

The soil moisture and water level sensors transmitted data directly to Thinger.io, thus enabling real-time visualization and storage on the platform.

Dashboards and monitoring

Custom dashboards showcased sensor data, including battery voltage and readings, enabling continuous monitoring and analysis.

Error handling and optimization

The platform monitored and managed connectivity issues as well as data transmission, ensuring efficient sensor operation even in remote areas.

What the data looks like

Three figures of the study, from the calibration of the probes to how closely the two probes of a node agree with each other over time.

Four scatter plots of moisture against sensor reading with a fitted decreasing curve: a single sensor (H51) for each probe on top, all sensors for each probe below
Calibration curves of a single sensor (top) and all sensors (bottom) for two probes: probe 1 (a and c) and probe 2 (b and d).
Five line charts: moisture of sensor 1, sensor 2 and their average over two periods, the range between both probes' records over the same periods, and the averaged record against the relative difference between probes
Moisture evolution over two periods (a, b), probe measurement ranges (c, d), and averaged record with relative difference (e).
Line chart of the coefficient of correlation between averaged moisture and relative difference for sensors 1 to 21 in catchment 1 and catchment 2, mostly between 0.4 and 0.95
Linear coefficient of correlation between averaged moisture and the relative difference between probes' measurements.

Conclusions

A scalable, low-cost solution for hydrological studies

The study demonstrated the effectiveness of the custom-designed hydrological monitoring system using Thinger.io.

By combining NB-IoT connectivity, cloud infrastructure and efficient sensor design, the system delivered accurate, real-time data on soil moisture and water levels, thus improving data reliability. As a result, it improved monitoring efficiency. Remote management of sensors enhanced battery performance while ensuring continuous data transmission, even in remote areas.

This monitoring system offers a scalable and low-cost solution for hydrological studies, enabling better water resource management and environmental monitoring.

Read the full study

The sensors, the network and the Thinger.io integration, as published by the research teams of the two universities.

Read the publicationSee the case study