AI & IoT Based Flood Monitoring and System

Vansh Goel, Anshum Shukla, Vishal Kumar Pandey and Meena Kumari

Department of Computer Science & Engineering (Internet of Things), ABESIT, Ghaziabad‑201009, Uttar Pradesh, India

Keywords: Internet of Things (IoT), Artificial Intelligence, Time Series Analysis (ARIMA) Algorithm, LID Performance,

Flood Early Warning System.

Abstract: Floods are a global hazard to both life and property, in addition to the infrastructure in place, which are

enhanced by issues like urbanization, inadequate drainage, and climate change. Proper flood management and

warning systems can minimize such losses. Latest solutions that use Internet of Things, and machine learning,

make it possible to monitor risk assessment, and alerts, in real time. Some of the techniques used in flood

prediction, especially in high-risk urban areas, include the use of AI models and smart sensors. IoT-enabled

systems that are equipped with sensors to measure water levels, rainfall, and weather parameters feed into

cloud platforms for the precise prediction of flood risks and efficient communication of warnings to

stakeholders. The decision-support algorithms in predictive models like the Time Series Analysis ARIMA

improve the accuracy of the prediction of floods and enhance the efficient response to emergencies through

data analysis on meteorological as well as hydrological entities. These systems, in combination with Low-

Impact Development (LID) practices, resilient network connectivity can provide early flood warning

capacities, enable rapid decision making, and reduce the effect of floods on vulnerable groups of people.

Testing and applications within real-life environments establish significant potential for its acceptance widely

to enhance the effectiveness in managing urban flooding as a whole and disaster readiness on a global scale.

1 INTRODUCTION

Floods are one of the most common natural disasters

around the world, affecting millions of people's lives,

infrastructure, agricultural systems, and economies

with far-reaching and devastating effects. Thus, the

rise in frequency and intensity of floods has also

brought along apprehensions of their impact as the

population of the world and urbanization continue to

grow. Although urbanization acts as an impetus for

economic growth, it is also known to increase the

flood risk associated with the region when coupled

with poor infrastructure. Water absorption into the

ground decreases with impervious surfaces created by

roads, buildings, and pavements. This increased

vulnerability, combined with that of climate change,

causes unpredictable weather patterns and

increasingly intense rainfalls, placing the urban area

at high risks of extreme flooding. The techniques of

flood management call for approaches toward

managing flood risks across cities and communities

worldwide and are motivated to advance.

Traditionally, the application of flood management

techniques largely depends on historical records

regarding floods. These records could be invaluable

but hardly ever satisfy the actual demand for in real-

time risk mitigation, as observed today. It is because

of these very limitations of the traditional systems

that researchers and policymakers have used the

opportunities presented by these new technologies,

such as IoT and ML, for developing dynamic,

responsive, and data-driven flood management

systems. For example, this type of IoT technology can

track and connect millions of sensors, which allows

for real-time monitoring of key flood indicators in the

environment, such as water levels, rainfall, and

weather conditions. At the same time, ML algorithms

can go through this data to find patterns and make

predictions for early flood warning and informed

decision-making to mitigate damage to lives,

property, and infrastructure.

Flood prediction models, especially machine

learning and AI-based ones, are revolutionizing the

approach to flood management and rely on massive

datasets and computational power. Such models make

highly accurate forecasts of rainfall, river discharge,

and flood levels using predictive algorithms such as

Time Series Analysis and Auto-Regressive Integrated

834

Goel, V., Shukla, A., Pandey, V. K. and Kumari, M.

AI & IoT Based Flood Monitoring and System.

DOI: 10.5220/0013874200004919

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies (ICRDICCT‘25 2025) - Volume 1, pages

834-841

ISBN: 978-989-758-777-1

Moving Average (ARIMA). Such predictions allow

for earlier warnings, which give local governments

and disaster response agencies crucial time to activate

the response plan, alert communities, and deploy

necessary resources. Improvements in cloud

computing allow for data centralisation and

processing in near real time from sensors placed

around the globe, easy to scale and integrate across

and between regions and even perhaps nationally.

The IoT allows for continuous monitoring and

automatic data collection, integral aspects of

providing timely and accurately provided flood risk

assessments. IoT technology permits a converged

network of strategically located sensors in flood-

prone areas to collect data in real time, thereby

affording the possibility of comprehensive

understanding of local environmental changes. Thus,

sensors might send real-time data to the cloud

platforms regarding a multiplicity of parameters

including precipitation rates, the moisture content in

the soil, levels of river water, and atmospheric

pressure. The final approach in data management

comprises a core of designing and developing reliable

flood prediction models together with making

actionable information delivery to the stakeholders.

Cloud infrastructures are rustic by strengthening

these systems, whereby easy integration of data

comes along with analytics at scale. They have made

data from vast sensor devices possibly available for

storage and processing. This ensures that local

authorities and emergency responders are kept

informed in order to respond quickly to such events.

Cloud computing bridged the gap of collecting data

and the actual action of decision-making within a

centralized data ecosystem. That is how flood

warnings through various channels, for example,

mobile applications, public alert systems, and even

direct notification to emergency teams, may be

delivered. Also, such systems are designed to work

under tough conditions with resilient network

connectivity ensuring constant monitoring even in

adverse weather events.

DSS forms a very important component of

effective flood management. It uses predictive

models and algorithms to evaluate the risk of flooding

in real-time and provides recommendations for

action. Decision-support algorithms incorporate not

only meteorological and hydrological data but also

contextual information, including land use patterns,

population density, and drainage system capacity, to

present an accurate risk profile. Contextual

integration allows for adaptive responses, with DSS

able to suggest preemptive actions such as activating

flood gates, issuing evacuation alerts, or mobilizing

emergency services in high-risk scenarios. Such

systems are more effective in terms of increasing

precision and response velocity of the strategies to

boost community resilience against flood disaster.

A new emerging concept of Low-Impact

Development in flood management practice that

seeks to combine a natural landscape feature and

good sustainable urban planning practice against risk

through flooding. LID approaches utilize the

application of permeable pavements, rain gardens,

and green roofs as measures to enhance the

infiltration rate of water while minimizing the runoff

and thereby intensifying the natural capacity for

urban areas to absorb. With LID coupled with IoT and

ML-based flood prediction systems, mitigation of

immediate risks associated with flooding also fosters

a long-term aspect of the resilience of the city at large,

thereby relieving pressure on the drainage systems so

that cities may face even extreme weather conditions

with a sense of security.

Advancement in IoT, ML, and cloud computing

improves multiple aspects of flood prediction, but

challenges in widespread usage are immense. These

high-tech systems have technical, financial, and

operational constraints that may not be achieved in

low- and middle-income regions. The extensive

sensor networks may be very expensive to deploy and

maintain resilient connectivity, especially in remote

or under-resourced areas. Moreover, issues of data

privacy and cybersecurity also need to be taken into

account because sensitive information gathered from

smart cities can easily be accessed by unauthorized

individuals. International cooperation, public-private

partnerships, and policy support are essential factors

in developing sustainable, accessible, and inclusive

flood management solutions against these challenges.

Case studies and pilot projects worldwide have

demonstrated the effective implementation of IoT and

ML-based flood management systems in spite of

these challenges. For instance, Japan and the

Netherlands have developed very large-scale, IoT-

enabled networks for flood monitoring that decrease

flood-related losses and strengthen community

preparedness. Working with technology vendors,

cities in flood-risk regions have created effective

alerting systems that combine real-time sensor

information, machine-learning-based forecasts, and

decision-support algorithms to inform the authorities

and citizens about a potential risk. Such examples will

transform the future of flood management and set a

precedent for other cities to become flood-resilient.

Instead, the confluence of IoT, ML, and cloud

computing technologies is helping open a new era in

flood management defined more by proactive risk

AI & IoT Based Flood Monitoring and System

835

reduction and improved disaster preparedness. Real-

time monitoring, predictive analytics, and decision

support systems allow communities to better

anticipate and respond to floods. LID-based practices

in urban planning could make cities less prone to

flooding over time and bring sustainable and resilient

development. As research continues, new

technologies are developed and policy frameworks

support this transition, the global adaptation of IoT

and ML-enabled flood management systems will

provide an opportunity for a safer, more resilient

future against the challenges of climate change.

1.1 Motivation

This paper synthesizes recent developments in AI and

IoT-based flood monitoring and rescue systems

aiming at pinpointing existing research gaps,

showcasing real-life applications, and offering

interdisciplinary insights. Focusing on floods as a

global challenge, it emphasizes the necessity of a

collaborative effort in order to resolve the issue.

Concerning the potential of AI and IoT technologies

to increase flood resilience and mitigate disaster

impacts, it speculates on future directions that provide

valuable perspectives for both academic discourse

and practical deployment in order to contribute to

enhancing global flood management strategies.

1.2 Objective

This is an advanced proactive system for monitoring

floods supported by AI and IoT technologies. The

system brings about mitigation of impacts from

floods. Sensors deployed to establish the water level

and early detection of danger, with support from AI

models, will help prepare for proper planning on how

to evacuate in time. Strong alert mechanisms with

SMS, mobile applications, and social media

notifications will ensure that warnings are brought

widely.

With location tracking and real-time

communication tools for emergency response teams,

rescue operations flow efficiently. Community

involvement is enabled through education and

resource availability. This system is scalable and

adaptable using open-source technologies and

resource optimization, which makes it affordable to

deploy in most geographical areas and flood

scenarios. The ultimate goal of this particular system

is to reduce loss of life and property and help develop

community resilience and disaster management

capacity.

2 LITERATURE SURVEY

Anisha Daniel P J et.al, 2021 proposed a model

Internet of Things (IoT) Based Monitoring of Floods

and Alerting System: Causes and implications of

flooding, the importance of flood monitoring, and

alert mechanisms are explained by the writer. He

recommends an Arduino Mega with water and rain

sensors, which would help predict flooding and hence

alert the authorities and other neighboring areas using

IoT technology. Real-time information and prediction

of the number of days before a given region gets

flooded will ensure prompt alarming to affected

regions. The IoT-based flood detection system for

early warning will provide timely alerts on level of

water and hence timely evacuations, preserving lives

and properties.

The project utilizes Raspberry Pi, LED, Buzzer,

ultrasonic sensor, and an Android Application for

monitoring floods and prediction. The model also

employs historical data and machine learning

algorithms for precise prediction of the flooding. This

means that users and authorities can conveniently

respond and manage a disaster. Himanshu Rai Goyal

et.al, 2021 proposed a post-management system with

the help of IoT devices and an AI approach. This

system uses the ANN algorithms to process data from

drainage condition sensors, rainfall sensors, and

network monitoring sensors. Although efficient, it

has limited scalability for extensive urban areas,

integration of various data sources, and validation

against real-time, large-scale floods. Other

comparisons with some other advanced AI models

remain unexplored, such as DNN and reinforcement

learning. Muhammad Izzat Zakaria el.al, 2023

reported a LoRaWAN-based IoT system using LoRa

Shield and ultrasonic sensors to monitor and alert of

floods in catchment areas. The system adopted

ARIMA for time series forecasting; however, it does

not attempt any adaptive spreading factors for

different scenarios, alternative communication

technologies, mesh networking towards the

expansion of coverage, and advanced data analysis

methods such as machine learning that enhance the

precision of flood prediction.

Kavitha Chaduvula et.al, 2021 presented an IoT-

based flood alert monitoring system using

microcontroller 8051 with pressure, water level,

temperature, and rainfall sensors. The proposed

system comprised the use of ARIMA and ANN

models for predictive analysis. Long-term reliability

and performance, integration with existing systems,

scalability for broad deployment, or energy

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

836

optimization strategies for remote use were not

addressed in the study.

Saeed Javanmaedi et.al, 2024 proposed an M-RL:

A Mobility and Impersonation Aware IDS for DDoS

UDP Flooding Attacks in IoT-Fog Networks, in

which they tackled the challenges of detecting and

mitigating DDoS attacks in IoT-fog environments.

The authors proposed a hybrid model, M-RL, which

fuses an RL technique with the RSS method for the

improvement of security in networks. The research

study finds out the lacunae present in today's anomaly

detection techniques and discusses the requirement of

comprehensive, energy-efficient solutions for power

nodes using fuzzy logic and operating with energy-

harvesting devices. In this study, Ismail Essamlali

el.al, 2024 has proposed a new architecture of Low

Impact Development (LID) stormwater management

using IoT and machine learning. By sensing rainfall,

soil moisture, water level, flow rate, and water

quality, the system enhances predictability and

develops decision-making capabilities. They point

out that no dedicated monitoring facilities and policy

framework exist for LID, and future work is in place

to overcome regulatory challenges and encourage

stakeholder collaborations.

Waheb A. Jabbar et al., 2023 presented an IoT-

based system using LoRaWAN for rural water quality

monitoring where pH, turbidity, temperature, and

TDS sensors have been integrated along with solar

cell power. Deep learning models enhance the data

handling process; however, the calibration process of

the system in a remote location is not perfect. Energy

consumption in idle and active states along with

integration of AI/ML for proactive monitoring and

more research work is required for effective

deployment.

Tae Sung Cheong et.al., 2024 proposed a flood

early warning system for small streams based on real-

time hydrodynamic information. Their approach

employs CCTV-based automatic discharge

measurement, water surface elevation gauges, and

surface imaging velocimetry, enhanced by the Robust

Constrained Nonlinear Optimization Algorithm

(RCNOA). Future work will delve into diverse runoff

forecasting within numerical models for rainfall

prediction, evaluating the effects of urbanization,

climate change, and land use on flood risk. Briefly,

Jinping Liu et al., 2022 reviewed the current state of

technologies for monitoring and forecasting urban

floods in the TC region through various types of

sensors, including satellite, radar, crowdsourced, and

IoT-based mobile sensors. Their finding brings out

the seriousness of the need to integrate advanced

sensor data into high-performance big data

techniques together with AI algorithms to be capable

of improving complex ones in urban flood forecasting

with intricate land and sewage systems. Martin Pies

et.al., 2024 elaborate advanced applications of IoT-

based wireless sensors for remote geotechnical

monitoring and structural diagnostics. Integration of

data accuracy along with outdoor antennas is the

primary aspect of their research, essentially using

geotechnical water-resistance sensors. Optimization

algorithms and signal prediction models have been

deployed; however, there is further scope for higher

sensor power efficiency, better water resistance, and

improved wireless signal transmission conditions,

mainly when underwater in different conditions.

2.1 Used Sensors

Table 1 shows the various sensors used and their

specifications.

Table 1: Sensor Specifications: Range and Accuracy.

2.2 Used Machine Learning Algorithm

Table 2: Algorithm Used.

Algorithm

Range

Accuracy

Time Series

Analysis

(ARIMA)

Short to

Medium-Term

Forecasting

Single-Variable

Forecasting

Dependence on

Historical Data

Quality

Trend and

Seasonality

Handling

Real-Time Data

Integration

Sensor

Range

Accuracy

Rainfall

Sensor

0 to 500 mm/h

Typically ±1% to

±5% of reading

Air Quality

Sensor

CO2 Levels: 0

to 5000 ppm

VOC (Volatile

Organic

Compounds): 0

to 1000 ppm

±(5-10)% of the

reading

depending on the

pollutant type

Flow Rate

Sensor

0.1 to 2500

L/min

±1% to ±3% of

reading

Temperature

& Humidity

Sensor

Temperature

Range: -40°C

to 125°C

Humidity

Range: 0% to

100% RH

Temperature:

±0.1°C to ±2°C

Humidity: ±2%

to ±5% RH

AI & IoT Based Flood Monitoring and System

837

Table 2 gives the range and accuracy of the algorithm

used.

3 PROBLEM STATEMENT

Worldwide, the frequency and intensity of flooding

events continue to rise as a result of climate change,

greater urbanization and changing weather patterns.

The outcomes of such flooding are severe, leading to

the loss of life and property in communities,

destruction or inundation of economic assets,

disruptions of port operations which facilitate

efficient trading activities amongst nations

precipitating into regional economies collapse as well

affecting biomes stability for example coral reef

susceptibility due sea temperature rise. Conventional

flood monitoring systems exist, but that requires data

to be gathered manually and they depend on localized

sensors which do not provide immediate assessment

of an extensive area due to the fact areas are so varied.

These limitations lead to slower response times,

insufficient early warnings and a greater dependency

on reactive rather than proactive disaster management

approaches.

In existing monitoring systems often only few

parts of the data are monitored which can make it

impossible to produce full, as fresh, insights that

would have been required for successful flood

prediction/migration. However, these systems

frequently lack the ability to offer forecasts on time

and at a high level of resolution while other

implementations that are based in resource-deprived

locations (such as remote rural regions) do not have

sufficient resources from sensor deployment and

maintenance. It is a deficiency that leaves wide open

the possibility of avoidable losses in affected areas.

The integration of Artificial Intelligence (AI) and

the Internet of Things (IoT) presents a transformative

solution to these challenges. By deploying IoT-

enabled sensors across critical flood-prone zones,

data on rainfall, river levels, soil moisture, and other

key variables can be continuously gathered in real

time. AI algorithms can then analyze this data to

produce predictive models, providing advanced flood

alerts and actionable insights for communities and

first responders. These systems allow for enhanced

situational awareness, early warning signals, and

automated response mechanisms, empowering

communities to take preventive action, optimize

resource allocation, and reduce potential damages. By

addressing the limitations of traditional monitoring,

an AI- and IoT-based flood monitoring system can

play a crucial role in adapting to the evolving climate

landscape and improving resilience against flood-

related disasters.

4 PROPOSED METHODOLOGY

The goal of this flood detection and alerting system is

to enable real-time monitoring, prediction, and alert

distribution through a centralized web portal. The

system integrates IoT sensors with a cloud-based

platform to collect, analyze, and disseminate

information, allowing users from various sectors such

as emergency response, healthcare, and the general

public to receive timely alerts and make informed

decisions. Figure 1 shows the block diagram of the

proposed system.

Figure 1: Block Diagram.

1. IoT Devices and Data Collection: The IoT

setup uses multiple sensors connected to an

Arduino Uno, which serves as the control unit.

The sensors continuously monitor

environmental parameters relevant to flooding.

Data collected by these sensors are transmitted

to a web server over the cloud using a Wi-Fi

module, allowing real-time access and storage

of information.

2. Data Storage and Cloud Integration: The

cloud serves as a centralized storage system,

holding both real-time and historical flood data

in CSV format. This data is accessible for

predictive analysis, enabling the detection of

potential flooding trends and making it available

for use in future machine learning models.

3. Real-Time Alerts and User Access: Based on

real-time data, alerts are triggered using visual

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

838

(LEDs), auditory (buzzers), and mobile app

notifications. Alerts are prioritized to notify

users according to the urgency, ensuring that

stakeholders can respond effectively to mitigate

flood damage.

4. Web Portal and Inter-User Communication:

The web portal connects users, allowing them to

access a range of services such as blood banks,

ambulance services, and hospitals that are

essential during flood emergencies. The portal

fosters communication and coordination among

users, enhancing emergency response

efficiency.

4.1 Role of Sensors

1. Ultrasonic Sensor: This sensor (figure 2)

measures the distance between the water surface

and a fixed point. In this project, it monitors

water level rise in real-time. By emitting

ultrasonic waves and calculating the time it

takes for the waves to bounce back, the sensor

estimates the water level, which helps detect

rising water before a flood occurs.

Figure 2: Ultrasonic Sensor.

2. Water Flow Sensor: A water flow sensor

(figure 3) detects the movement and rate of

water flow. This information helps identify

potential blockages or changes in water flow,

which can contribute to flooding. Monitoring

water flow rates enables the system to respond

proactively if flow patterns indicate flood risks.

Figure 3: Water Flow Sensor.

3. Water Level Sensor: The water level sensor

(figure 4) directly measures the water level in

reservoirs, rivers, or other bodies of water. The

data helps determine if the water level is

reaching critical points. This sensor is essential

for confirming that the water has exceeded safe

thresholds, signaling the need for immediate

alerts.

Figure 4: Water Level Sensor.

4. Temperature and Humidity Sensor:

Temperature and humidity sensors (figure 5)

monitor environmental factors that could impact

flooding conditions. High humidity and

significant temperature variations often

correlate with heavy rainfall, which can lead to

floods. Tracking these parameters enables better

AI & IoT Based Flood Monitoring and System

839

prediction of weather patterns that may result in

flooding.

Figure 5: Temperature and Humidity Sensor.

4.2 Predictive Analysis with the

ARIMA Algorithm

The AutoRegressive Integrated Moving Average

(ARIMA) algorithm is a statistical model used for

forecasting based on historical time-series data. In

this project, ARIMA analyzes historical flood data

(stored in CSV format) to predict future flood events.

The model processes trends and seasonal patterns in

the data, enabling it to forecast rising water levels or

other flood indicators. By integrating ARIMA

predictions with real-time sensor data, the system can

provide advanced alerts and help authorities take

preventive actions before flooding reaches critical

levels.

5 RESULT

All the proposed distributed flood detection and

alerting systems use a single centralized web portal to

monitor floods in real time, predict them, and,

consequently, warn the people. A cloud platform

integrates IoT sensors for the real-time collection and

analysis of environmental parameters. The

stakeholders who should receive an alert in time

before any flooding occurs in the area in which they

operate are the emergency responders, health-care

providers, and the public. The IoT setup has sensors

connected to an Arduino Uno that do level, flow rate,

and temp and humidity monitoring. Data is

transmitted in real time with a Glowlink Wi-Fi

module to a cloud server hence available and stored

continuously. If the data which are stored in the Cloud

in CSV format is helpful in business forecast analysis,

let's leave it as it is for later machine learning use

cases.

Figure 6: Circuit Diagram.

The urgency of the potential flood threats that

arrive is notified as push LEDs and buzzers and

mobile through alerts in a semi-automated

synchronized process, thus enabling people to take

any action on them quite early. The web portal will

offer access to needed services and facilitate a

medium for the communication between users which

could eventually come in handy as an emergent

response.

While the ARIMA algorithm processes historical

data for predictive analysis, it identifies changing

trends patterns including those in flood events. The

system augments predictions of ARIMA with live

sensor inputs to allow for proactive alerting that

equips the authorities with the information required in

order to take preventive actions before the flood

conditions aggravate. Figure 6 depicts the circuit

diagram of the proposed method.

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

840

6 CONCLUSION AND FUTURE

SCOPE

This project will show the possibility of creating an

alert system that would mitigate the risks of flooding.

By this, it could monitor what data the sensors have

held, and there is a possibility of the incorporation of

sensors within the system for an accurate and efficient

system of flood detection. It may prove to benefit

other government agencies or authorities, whereby

this benefits society in the management and

mitigation of flooding, a hazard natural disaster. It

gives wide monitoring of all factors that could cause

flooding. When water rises and its speed increases, it

gives an alert immediately. In simple words, this

system gives improved access within the

management and response to this catastrophic

occurrence. It aids the community to take informed

decisions and plan how to overcome this natural

catastrophe effectively.

This innovation shall decrease flooding which

essentially encompasses an ultrasonic sensor

measuring the flood levels on the road, a live

streaming camera to monitor live for the occurrence

of flooding, and Serial Communication that sends text

messages with warnings whereby date, time, water

level, and accessibility status are listed. There are

three modules under the system: Users, Logs, and

Contact Numbers. These can be edited by the admin.

For precision, the sensor unit is suggested to be

installed in front of the system, perpendicular to the

floodwater, and attached with a pole, 3 to 3.5 meters

long. For continuous operation, the flood sensors and

microcontrollers are all powered with an 80,000

Ampere-Hour (mAh) Solar Power Bank that ensures

uninterrupted, round-the-clock water level detection

and network data transmission.

REFERENCES

Anisha Daniel P J, Abhishek M L, Frelbin Nazeer3 and Ann

Johny “Internet of Things (IoT) Flood Monitoring and

Alerting System”, International Journal of Innovative

Research in Technology, Vol.8, Dec.2021.

Himanshu Rai Goyal, Kamal Kumar Ghanshala, Sachin

Sharma “Post flood management system based on smart

IoT devices using AI approach”, Materials Today:

Proceedings, Jan.2021.

Ismail Essamlali, Hasna Nhaila, Mohamed El Khaili “A

new architecture of Low Impact Development (LID)-

based stormwater management system through Internet

of Things (IoT) and Machine Learning integration”,

Case Studies in Chemical and Environmental

Engineering, Vol.10, Sep.2024.

Jinping Liu, Hyo-Seob Cho, Sazali Osman, Hyeon-Gyo

Jeong, Kwonmin Lee “Review of the status of urban

flood monitoring and forecasting in TC region”,

Tropical Cyclone Research and Review, Vol.11, July

2022.

Kavitha Chaduvula, Kranthi kumar k., Babu Rao

Markapudi, Rathna Jyothi Ch. “Design and

Implementation of IoT based flood monitoring system

using microcontroller 8051”, Materials Today:

Proceedings, 2021.

Martin Pies, Jan Velicka, Radovan Hajovsky “Advanced

IoT-Based Wireless Sensors for Remote Geotechnical

Monitoring and structural Diagnostics”, IFAC-

PapersOnLine, Vol.58, 2024.

Muhammad Izzat Zakaria, Waheb A, Jabbar, Noorazliza

Sulaiman “Development of a smart sensing unit for

LoRaWAN-based IoT flood monitoring and warning

system in catchment areas”, Internet of Things and

Cyber-Physical Systems, Vol.3, May 2023.

Saeed Javanmardi, Meysam Ghahramani, Mohammad

Shojafar, Mamoun Alazab, Antonio M. Caruso “M-RL:

A mobility and impersonation-aware IDS for DDos

UDP flooding attacks in IoT-Fog Network”, Computers

& Security, Vol.140, Feb.2024

Tae Sung Cheong, Seojun Kim, Kang Min Koo

“Development of measured hydrodynamic

information-based flood early warning system for small

streams” , Water Research, Vol.263, Aug 2024.

Waheb A. Jabbar, Tan Mei Ting, M. Fikri I. Hamidun,

Ajwad H. Che Kamarudin, Wenyan Wu, Jamil Sultan,

AbdulRehman A. Alsewari, Mohammed A.H. Ali

“Development of LoRaWAN-based IoT system for

water quality monitoring in rural areas ", Expert

Systems With Applications, Vol.242, Dec.2023.

AI & IoT Based Flood Monitoring and System

841