Research Overview

Research Background: What Problem Did This Study Address?

Landslides are among the most destructive geological hazards, especially in mountainous and high-rainfall regions. They can damage roads, buildings, utilities, and communities, and they are often triggered by rainfall-induced changes inside the slope. When rainwater infiltrates the soil, it changes soil moisture, matric suction, pore-water pressure, and suction stress. These changes can reduce slope stability and eventually lead to shallow landslides.

Wireless Sensor Networks, or WSNs, are increasingly used for landslide monitoring because they can collect field data from multiple locations and transmit them to a base station or cloud server. However, landslide monitoring WSNs face practical challenges. Sensor nodes are often deployed in remote or difficult terrain, where replacing batteries, repairing communication links, or manually collecting data is expensive. Energy consumption, packet loss, data delay, routing efficiency, and network reliability therefore become central issues.

This study proposed an adaptive landslide monitoring system based on Fuzzy Logic-based Particle Swarm Optimization, abbreviated as FLPSO. The method was designed to improve multi-mobile-agent itinerary planning in WSNs, reduce energy consumption, and improve network performance. A case study was conducted in Shiradi village near Mangalore, India, a high-rainfall area with landslide susceptibility.

UbiBot RS485 sensors were used as part of the field monitoring setup to track soil moisture-related variation during rainfall infiltration. These data were combined with matric suction, rainfall, tilt, and other sensor data to support slope-stability and factor-of-safety analysis.

The Specific Role of UbiBot in the Study

In this study, UbiBot RS485 sensors were used as part of the field sensor deployment in the landslide monitoring system. The paper does not describe UbiBot as independently predicting landslides or proving the effectiveness of the FLPSO algorithm. Instead, UbiBot sensors contributed field measurements related to soil water variation, which formed part of the geotechnical monitoring dataset used to evaluate rainfall-induced slope behavior.

The deployment was carried out in the Shiradi Ghats / Shiradi village region near Mangalore, India. This area is described as prone to rainfall-induced shallow landslides, with soil strata commonly within about 2 metres of the natural slope surface. Because shallow landslides in this region may occur at depths of less than 1 metre, the monitoring system needed to observe changes at different soil depths.

The paper states that sensor nodes were installed at 0.5 m, 1.0 m, and 1.5 m below the ground surface. UbiBot RS485 sensors were installed to monitor variation in volumetric water level / volumetric water content as precipitation infiltrated downward into the topsoil layers. As rainfall entered the soil, the wetting front moved through the slope, changing saturation and moisture conditions. These changes are important because they affect matric suction and suction stress, which in turn influence slope stability.

The UbiBot data were not used in isolation. They were part of a multi-sensor field monitoring architecture. MPS-8 sensors were installed to measure matric suction. LM31 tilt meters were used to detect slope movement. WTB100 tipping-bucket rain gauge sensors measured rainfall amount and intensity. MF4003 sensors were installed to identify and analyse debris-flow velocities. The monitored data were transmitted through a wireless sensor network architecture involving slave nodes, master nodes, sink nodes, 4G communication, a cloud server, and a web service.

The data contributed by UbiBot sensors supported environmental and geotechnical monitoring. Specifically, soil moisture / volumetric water variation helped describe rainfall infiltration and wetting-front movement. Combined with matric suction data, these measurements allowed the researchers to calculate suction stress and then estimate the factor of safety for infinite-slope stability analysis.

Therefore, UbiBot’s role in the study was field data acquisition for rainfall-induced slope monitoring. It provided soil water-state information that helped the research team link rainfall, moisture infiltration, suction stress, and slope stability within the broader FLPSO-based WSN monitoring framework.

Research Methods and Data Collection Approach

The study combined field deployment, geotechnical analysis, wireless sensor network design, and algorithmic simulation.

First, the researchers developed a landslide monitoring system architecture based on wireless sensor networks. The system included slave nodes, master nodes, sink devices, base stations, mobile agents, a cloud server, and a management platform. Slave nodes collected data from sensors installed at landslide observation points. Master and sink nodes handled data collection, protocol control, communication, and transfer to the cloud server.

Second, the field deployment was conducted in Shiradi village / Shiradi Ghats near Mangalore, India. The region was selected because of high annual rainfall, changing climatic conditions, and landslide susceptibility. Sensors were installed at 0.5 m, 1.0 m, and 1.5 m below the ground surface to observe variations in soil conditions at multiple depths.

Third, multiple sensing devices were used. UbiBot RS485 sensors monitored changes in soil water state during rainfall infiltration. MPS-8 sensors measured matric suction. LM31 sensors measured slope movement or tilt. WTB100 tipping-bucket rain gauges measured rainfall quantity and intensity. MF4003 sensors were used to detect debris-flow velocity through wire-break digital signalling.

Fourth, soil samples from the study area were analysed to determine geotechnical properties. The paper reports parameters such as saturated volumetric water content, residual volumetric water content, dry unit weight, effective cohesion, internal friction angle, percent fines, and liquid limit. These parameters were used in slope-stability calculations.

Fifth, the researchers calculated suction stress and factor of safety using recorded monitoring data. Rainfall-induced changes in soil suction were used to estimate the factor of safety at depths such as 0.5 m and 1 m.

Finally, the FLPSO algorithm was evaluated in MATLAB R2019a. The simulation used a three-dimensional WSN landscape with network parameters including number of deployed nodes, source nodes, node energy, transmission range, mobile-agent code size, and aggregation ratio. FLPSO was benchmarked against AEEF, LDCSPC, LDMWSN, and EEELDS using packet delivery ratio, energy delay product, packet loss ratio, task energy, and throughput.

Key Research Findings

The field analysis showed that rainfall affected soil suction and slope stability in the Shiradi monitoring area. The paper reports that suction stress changed between about 0.2 and 2 kPa according to rainfall conditions. Using recorded data and soil parameters, the factor of safety was calculated over a one-year period.

At a depth of 0.5 m, the factor of safety fluctuated from approximately 1.7 to 1.4. At a depth of 1 m, it fluctuated from approximately 1.45 to 1.32. Because no landslide occurred during the monitoring period, the factor of safety did not drop below 1.0. This supported the use of real-time rainfall and soil-condition monitoring for slope-stability assessment and threshold-based warning design.

The FLPSO method improved wireless sensor network performance compared with benchmark methods. The study reports that FLPSO improved packet delivery ratio by about 14.15% compared with EEELDS, and by larger margins compared with LDMWSN, AEEF, and LDCSPC. Energy delay product improved by about 11.15% compared with AEEF. Packet loss ratio improved by about 10.15% compared with AEEF. Task energy consumption was reduced compared with AEEF, EEELDS, LDCSPC, and LDMWSN. Throughput increased by about 20.1% compared with AEEF, and by larger margins compared with the other methods.

The statistical analysis also supported the reported performance advantage. The paper used a Friedman test at a 5% confidence level and reported p-values below 0.05 for all parameters. The authors concluded that FLPSO performed better than the compared methods across the evaluated metrics.

Overall, the study found that combining field monitoring with FLPSO-based multi-mobile-agent itinerary planning can improve both landslide-risk monitoring and WSN energy efficiency.

What This Means for Landslide Monitoring Applications

This study shows that landslide monitoring requires both geotechnical sensing and communication-network optimisation. Measuring rainfall alone is not enough. For rainfall-induced shallow landslides, it is important to monitor how water infiltrates the soil, how matric suction changes, and how these changes affect the factor of safety.

For field monitoring, UbiBot RS485 sensors contributed soil moisture-related information that helped describe the wetting process in the slope. When combined with matric suction data, rainfall data, tilt data, and soil properties, this information supported a more physically meaningful assessment of slope stability.

For wireless sensor network design, the study highlights the importance of energy-efficient routing. Landslide monitoring networks are often deployed in remote terrain, where long-term operation is critical. FLPSO was proposed to select more efficient routes for mobile-agent-based data collection, reducing energy consumption and improving data delivery performance.

For early warning systems, the approach suggests that real-time monitoring data can be converted into safety-factor estimates and warning thresholds. If future monitoring detects a factor of safety approaching critical levels, the system could support earlier alerts for evacuation and mitigation.

Application Value of UbiBot Devices

In this study, UbiBot RS485 sensors demonstrated application value as field sensing components in a landslide monitoring network.

First, they supported soil water-state monitoring during rainfall infiltration. This is important because shallow landslides are strongly linked to changes in moisture, saturation, suction stress, and wetting-front movement.

Second, the sensors were deployed at multiple depths together with other instruments. By collecting data below the ground surface at 0.5 m, 1.0 m, and 1.5 m depth levels, the monitoring system could observe how rainfall affected different soil layers.

Third, UbiBot data complemented matric suction measurement. Moisture variation and matric suction together helped the researchers estimate suction stress and factor of safety.

Fourth, the sensor data became part of a wireless sensor network workflow. The field data were transmitted through slave nodes, master nodes, sink nodes, 4G communication, and cloud-server infrastructure, supporting remote monitoring and analysis.

Fifth, UbiBot’s role illustrates how modular RS485 sensing can be incorporated into broader landslide early warning systems. The device was not the algorithm itself, but it supplied field environmental data needed for geotechnical analysis and network-based warning logic.

Extended Application Scenarios

The monitoring approach discussed in this study can be extended to several related scenarios:

1. Rainfall-induced landslide monitoring
   Used to monitor soil moisture variation, matric suction, rainfall, tilt, and slope movement in high-risk slopes.
2. Highway and railway slope safety
   Used along transportation corridors where shallow landslides may damage roads, tracks, embankments, or cut slopes.
3. Mountain village early warning systems
   Used to provide field data for local landslide warning platforms in rainfall-prone settlements.
4. Wireless sensor network research
   Used as part of WSN deployments that evaluate energy-efficient routing, mobile-agent planning, and cloud-based monitoring.
5. Soil infiltration studies
   Used to record moisture changes at different depths during rainfall events.
6. Unsaturated soil mechanics research
   Used with suction sensors to study how moisture and matric suction affect slope stability.
7. Disaster management platforms
   Used to feed real-time field monitoring data into alert dashboards and emergency response workflows.
8. Long-term slope health monitoring
   Used for continuous observation of seasonal and rainfall-driven changes in soil water conditions.

FAQ

1. Which UbiBot product was used in the study?

The paper mentions UbiBot RS485 sensors.

2. What data did UbiBot collect?

UbiBot RS485 sensors were used to monitor changes in volumetric water level / volumetric water content as rainfall infiltrated into the topsoil layers.

3. Where were the sensors deployed?

The monitoring system was deployed in Shiradi village / Shiradi Ghats near Mangalore, India, a landslide-prone area with high rainfall.

4. At what depths were sensors installed?

The paper states that nodes were set at 0.5 m, 1.0 m, and 1.5 m below the ground surface.

5. What was the sampling frequency?

The paper does not explicitly state the UbiBot RS485 sampling frequency.

6. How long did the study last?

The paper describes one year of monitoring and analysis, with factor-of-safety variation shown over a one-year period beginning in 2021.

7. Did UbiBot predict landslides by itself?

No. UbiBot sensors collected soil moisture-related data. The research team combined these data with matric suction, rainfall, tilt, soil-property data, and FLPSO-based WSN routing to support landslide monitoring and prediction analysis.

8. What other sensors were used?

The system also used MPS-8 sensors for matric suction, LM31 sensors for slope movement, WTB100 tipping-bucket rain gauges for rainfall, and MF4003 sensors for debris-flow velocity detection.

9. What was the main algorithmic contribution?

The main contribution was an FLPSO-based multi-mobile-agent itinerary planning method for wireless sensor networks, designed to reduce energy consumption and improve data transmission performance.

10. What was the main result?

FLPSO improved network metrics such as packet delivery ratio, energy delay product, packet loss ratio, task energy, and throughput compared with benchmark methods, while field monitoring supported factor-of-safety analysis for rainfall-induced slope instability.