1. Introduction

The Indian government started the smart city initiative to raise living standards in major cities. As various networks (e.g., LAN, WLAN, and 3GPP) are becoming increasingly interconnected to support the smart city infrastructure, proper underground drainage systems must be implemented, which is a major goal of the initiative. Cities must be clean, and the spread of waterborne diseases must be stopped. Many cities in India have weak systems for monitoring underground sewage, which has led to problems, such as the contamination of drinking water and the spread of disease. The use of open-channel drainage systems, which is common in rural areas and urban areas with less development, has particular problems because of the need for regular maintenance and the possibility of blockages in the systems. They also presented a health risk to the nearby population because they serve as a breeding ground for parasites and mosquitoes that spread disease. On the other hand, closed drainage systems are safer and more hygienic when used in urban and metropolitan areas, but they present maintenance and cleaning challenges. Cleaning closed drainage systems is laborious and time-consuming, and workers are exposed to serious health risks because of the need to enter access chambers that collect poisonous gases. Furthermore, unanticipated obstructions in drainage lanes can result in the flooding of drainage water on roads, contaminating clean water and causing the spread of dangerous waterborne diseases.

This paper suggests the development of smart underground infrastructure with sensor networks to detect various drainage parameters, including the precise location of blockages and images of blockages, to address these issues. The proposed system requires a crusher system to clear obstructions and a crane system to lift the trash into the access chamber. In addition, various sensor networks are required to detect toxic gas detection systems, access chamber status indicators, and water flow, viscosity, and depth. Therefore, improving the upkeep and cleaning of underground drainage systems in metropolitan and urban cities requires the development of an automated drainage monitoring system. An automated drainage monitoring system is required to maintain and clean urban underground drainage systems.

2. Related Work

WSNs are a cost-effective way of monitoring the environment. WSNs are battery-powered sensor nodes with processors, communication devices, and sensors. These nodes in the monitored area communicate to gather and process data, which is sent to a base station in the center for analysis. WSNs are used for environmental, precision agriculture, structure health, and underground monitoring. This literature review focuses on underground wireless sensor network opportunities and challenges.

Prasad Reddy et al. proposed a magnetic induction-based wireless underground sensor network ^[1]. The authors described an underground magnetic induction-based communication system. The 25-centimeter-range device has a transmitting and receiving unit. The results show that the proposed underground communication system is reliable and efficient. Kim et al. proposed a wireless subterranean sensor network using magnetic core antennas and multiple surface acoustic wave sensor modules ^[2]. The authors described a system with many surface acoustic wave sensor modules for sensing and magnetic core antennas for communication. The system under consideration in a laboratory can detect and categorize diverse entities. Wang et al. ^[3] proposed a dynamically uncertain and wave-disturbed AUV path planning and following control system. The AUVs were planned and controlled using models in uncertain and disturbed environments. The simulations tested the system. Despite the uncertainties and disturbances, the system planned and controlled paths accurately in the experiment. Stewart et al.'s dynamic modeling system ^[4] has improved AAUV-specific passive drainage.

Manzanilla et al. suggested an autonomous navigation system for unmanned underwater vehicles (UUVs) ^[5]. The authors proposed a vision-based UUV navigation and control system. The simulations suggested that the system under consideration could accurately navigate and manage unmanned underwater vehicles (UUVs). Wang et al. ^[6] developed an organization that could detect multiple drainage pipeline blockages by combining support vector machine features with VMD characteristics. The authors' drainage pipeline obstruction detection and removal system used machine learning. The proposed system successfully identified drainage pipeline obstructions using real-world data. Nielsen et al. ^[7] described the use of genetic programming to model an autonomous underwater vehicle fitted with a plastic grabber. The main aim of the experiment was to recover plastic items from the ocean. The researchers presented a methodology using an autonomous underwater vehicle simulation to retrieve plastic materials. Jawhar et al. in reference ^[8] proposed an architecture for using autonomous underwater vehicles in wireless sensor networks. The authors described an architecture for wireless sensor networks for pipeline monitoring that uses autonomous underwater vehicles as mobile sensor nodes. The proposed system was tested in a simulation environment. The results showed that it is capable of precisely monitoring underwater conditions.

Soung-YueLiew and ShenKhangTeoh ^[9] proposed a data store-and-delivery approach for air-ground collaborative wireless sensor networks. They developed a data storage and delivery framework that can deliver ground-based sensor data to an unmanned aerial vehicle (UAV) or a ground station while storing it. They conducted simulation experiments to assess their suggested strategy. Haswani and Deore ^[10] proposed a Web-based real-time underground drainage or sewage monitoring system using WSNs. They used a ZigBee-based WSN for gathering and transmitting real-time data from the underground drainage system to the control center. They used both simulation and in-person tests to assess their suggested system.

In wireless underground sensor networks (WUSNs), Hoang ThiHuyenTrang, Le The Dung, and SeongOun Hwang ^[11] examined the connectivity of underground sensors. They used mathematical modeling and simulations to assess the connectivity of various WUSN topologies. They also suggested a hybrid topology that could enhance the WUSN connectivity. Sahana et al. ^[12] reviewed underwater wireless sensor network (UWSN) routing protocols and their difficulties. They discussed various UWSN routing protocols and identified problems that needed to be solved to boost UWSN performance. Choudhary et al. ^[13] reviewed autonomous underwater vehicles (AUVs) underwater navigation systems. They identified the issues and future directions for research in this area while discussing various AUV navigation systems.

Moridia et al. ^[14] developed wireless sensor networks for monitoring and communication systems underground. They used a ZigBee-based WSN to monitor underground environments, such as mines and tunnels. They conducted experiments to gauge the effectiveness of their suggested system. Cao et al. ^[15] developed an unmanned surface vehicle (USV) to monitor water quality. Data on water quality was gathered using a multi-sensor system and sent to the command center. They conducted actual experiments to gauge the effectiveness of their suggested system. Wu et al. ^[16] proposed a long-range wide area network (LoRaWAN) for agricultural wireless underground sensor networks (WUSNs). They improved the coverage of WUSNs in agricultural fields and increased their communication range using LoRaWAN. They used simulations to assess their suggested system.

Mishra et al. proposed an underground coal mine IoT-based multimode sensing platform ^[17]. A multimode sensing platform measured the pressure, humidity, gas concentration, and temperature. Their proposed platform was tested using simulations. \"{U}nsal et al. proposed WSN power management for underground mining ^[18]. Dynamic voltage scaling reduced the underground mining WSN power consumption. They tested their plan with simulations. Akkas reviewed WUSNs for mines and miners’ safety ^[19]. Jalaja and Jacob ^[20] proposed a mobile-element-based adaptive data collection scheme for sparse underwater sensor networks (UWSNs). They discussed WUSN applications in mines and miners’ safety, challenges, and future research. Mobile elements with acoustic communication tools collected data from sparse network sensor nodes. Avoiding redundant data transmission would improve network coverage and save energy. The simulation showed that the proposed scheme outperformed other schemes in coverage and energy efficiency. Zhu et al. ^[21] suggested using the Gaussian mixture model (GMM) clustering and complete ensemble empirical mode decomposition (CEEMD) to detect underwater drainage pipe blockages. The system was designed to improve submerged drainage pipe blockage detection. The authors reported that their system outperformed others in accuracy and false alarm rates.

See et al. ^[22] suggested a Zigbee-based WSN for sewerage monitoring. The proposed system monitored sewerage conditions in real time and reduced maintenance costs by detecting blockages and leaks early. The experimental evaluation showed that their system was more reliable and energy efficient than others. Kennedy and Bedford studied underground wireless networking standards for tunneling and mining ^[23]. The authors evaluated and compared the throughput, dependability, and signal attenuation of the Ultra-Wideband (UWB), Wi-Fi, and Zigbee standards. Zigbee had the best signal attenuation and reliability.

Moridi et al. investigated a Zigbee radio wave attenuation-based underground monitoring and communication system ^[24]. The proposed system improved tunneling and underground mining communication. The authors reported that the system was reliable and low-power compared to other methods. Girisrinivas and Parthipan proposed an IoT-based DOMS ^[25]. For safety and health, the proposed system could monitor drainage overflow in real time. The experiments showed that the authors' system was more accurate and faster. Manual scavenging-cleaning septic tanks and sewer lines by hand-killed 631 people in India in the past decade, according to The Hindu ^[26]. Wireless sensor networks in underground monitoring and communication systems can reduce manual scavenging and improve worker safety and health in mining and sewerage.

This literature review concludes with a comprehensive review of wireless sensor networks for underground and underwater monitoring. The review papers covered data collection, network connectivity, communication technologies, and monitoring systems. The papers reviewed provide useful information and suggest promising research directions.

3. The Proposed Scheme

The methodology for the automated drainage monitoring system is based on a consideration of various parameters that need to be monitored to prevent issues that people face during the rainy season, contamination of drainage water with pure water, accidents involving drainage cleaning workers, the transmission of infectious diseases by mosquitoes, and accidents caused by open access chambers. The suggested system uses a sensor network to identify drainage parameters, such as water level, flow, pressure, and toxic gases. The system also features a access chamber status indicator.

The suggested model includes using an autonomous underwater vehicle (AUV) to address obstructions in the drainage system. The AUV can locate obstructions, take pictures of obstructions, destroy obstructions, and lift the crushed trash to the access chamber. The proposed smart system consists of the design of sensor nodes to gather various parameters, the connection of all sensor nodes using GPON with PoE port, the provision of underwater radio frequency coverage using a Remote Radio Head (RRH) or Remote Radio Unit (RRU), and the assembly of various sensors and equipment on the AUV to clear obstructions with a remote command.

Sensor Node (SN), Surface Node, Network Control Center (NCC), and Compilation of various sensors and equipment on the AUV comprise the four components of the proposed model. The SN uses different communication protocols to upload the data collected to the NCC, such as GSM, Wi-Max, Wi-Fi, satellite, and optical fibers (GPON with PoE). The SN performs various sensing operations, such as pressure, level of savage water, chemical compositions, and access chamber status indicators.

The NCC gathers information from surface nodes and locates drainage blockages underground. It issues orders telling the AUV to travel to the location of the obstruction, clear it, and monitor its operation. After receiving the command from the NCC, the AUV removes the obstruction, takes pictures with a night vision camera, and sprinkles disinfectant into the drainage system.

The AUV's RF transceiver, GPS, night vision camera, and sprinkler to clean the drainage are all factors that were considered when building it. Overall, the automated drainage monitoring system methodology offers an effective and efficient way to enhance the upkeep and cleaning of underground drainage systems and stop the spread of waterborne diseases in India's urban and metropolitan areas.

3.1 ADVMS Architecture

The following steps outline the suggested methodology for developing the Automatic Drain Vision Management System, or ADVMS:

1.Identification of Parameters: Consider various parameters that must be monitored to stop issues that people encounter during the rainy season, contamination of drainage water with pure water, misadventures involving workers who clean the drainage system, transmission of infectious diseases by mosquitoes, and accidents caused by open access chambers.

2.Sensor Network: Utilize a sensor network to identify drainage-related factors, such as water level, flow, pressure, and toxic gases.

3.Autonomous Underwater Vehicle: Employ an autonomous underwater vehicle (AUV) to clear up clogs in the drainage system. The AUV can locate obstructions, take pictures of obstructions, destroy obstructions, and lift the crushed trash to the access chamber.

4.Sensor Node: Develop sensor nodes to gather various parameters, such as pressure, savage water level, chemical compositions, and indicators of access chamber status.

5.Establish a surface node to gather information from sensor nodes.

6.Network Control Center: Gather information from nodes on the surface and locate drainage blockages underground. Order the AUV to travel to the obstruction's location, remove the obstruction, and then monitor the vehicle's performance.

7.AUV Compilation: Combine various sensors and equipment on the AUV to clear obstructions by receiving a remote command. To clean the drainage, consider using an RF transceiver, a GPS, a night vision camera, and a sprinkler.

Fig. 1 summarizes the ADVMS architecture, which offers a practical and efficient solution to enhance the upkeep and sanitation of underground drainage systems and stop the spread of waterborne diseases in India's urban and metropolitan areas. The Network Control Center can access the data that the Sensor Node uploads to it using communication protocols like GSM, Wi-Max, Wi-Fi, satellite, and optical fiber (GPON with PoE). The Sensor Node gathers various drainage system parameters. The Surface Node gathers data from the Sensor Nodes and uploads it to the Network Control Centre to locate blockages in the underground drainage system and instruct the autonomous underwater vehicle (AUV). The AUV uses a night vision camera to take pictures while clearing the obstruction and sprinkling disinfectant into the drainage system.

Fig. 1. ADVMS Architecture.

4. Performance Evaluation

The outcomes of the 3D model can offer useful perceptions of how well the drainage system functions. For example, the simulation might identify places where water is not moving as effectively as it should or places where it might pool and cause flooding. These problems can be identified before construction, allowing designers to alter the drainage system to guarantee it works correctly and efficiently. This can save time and money by avoiding expensive repairs and ensuring that the drainage system can handle the anticipated volume of water.

Overall, the 3D model offers an effective tool for designing and optimizing drainage systems, ensuring they are strong, effective, and resilient in the face of shifting weather patterns and other environmental factors. A hydrological model was constructed using the subsequent measurements: The object dimensions in question are 16 feet in length, two feet in width, and 1.5 feet in height. Access chamber-1, Access chamber-2, and Access chamber-3 are the three access chambers in the system. Fig. 2 presents the design of the drainage model. The remaining portion of the model was covered with transparent acrylic sheets to ensure proper drainage system monitoring. Sensor node access chamber caps, each measuring 2${\times}$2 feet, were installed on the access chambers. The caps had four sensor node slots to accommodate the various sensors. The gas sensor, ultrasonic contactless level sensor, and air pressure sensor were placed in the central slots, while the access chamber status indicator sensor was set aside for the left slot. The construction of the model made it possible to monitor the drainage system effectively and provided useful information about how well it was working.

Fig. 2. 3-D-Modeling of Drainage System.

Fig. 3 presents the actual drainage model. Fig. 4 depicts an MQ-4 gas sensor that finds methane gas (CH$_{4}$) in drainage systems. The MQ-4 sensor is made to find the primary component of natural gas, CH$_{4}$, in drainage systems. An analog output can represent the CH$_{4}$ concentration in the drainage by connecting the MQ-4 sensor to an Arduino Uno microcontroller.

Fig. 3. Prototype of drainage model.

Fig. 4. Simulation of the node for Gas sensing.

Fig. 5. IR Sensor based level sensing setup.

The sensor also has a digital output that alerts when methane levels reach or surpasses a predetermined threshold. This alert can turn on an external LED and notify users when the CH$_{4}$ gas levels in the environment may be dangerous. The HX710B atmospheric pressure sensor module was placed in the test using an Arduino microcontroller. The module has an altitude resolution of 10 cm and is intended for use in altimeters and variometers. A low-power, 24-bit ADC with internally calibrated coefficients and a pressure sensor with high linearity are both included in the sensor module. It provides accurate 24-bit digital pressure and temperature readings, and the user can choose from various operating modes to optimize conversion efficiency and power usage. Fig. 6 presents the pressure measurement results of the HX710B Atmospheric Pressure Sensor Module. The sensor module uses a high-precision AD sampling chip and an 0-40KPa air pressure sensor to detect air pressure and water level. It can be attached to a 2.5mm hose. The findings show that the module measures the atmospheric pressure with accuracy and dependability, making it suitable for various applications, including weather monitoring and altitude tracking. Using a GPON network, the drainage parameters from various sensor nodes were gathered. A PoE switch was used to connect each sensor node and supply power. An ultrasonic detector measures the distance of an object using sound waves.

Fig. 6. Pressure Sensor.

The node operates by sending out a sound wave at a particular frequency and then waiting for the wave to return.

The distance between the ultrasonic sensor and the object can be calculated by timing the emission and reception of the sound wave. The level of water can be determined using this kind of detector.

Fig. 5 shows the use of an ultrasonic sensor to measure the water level. The sensor sends out a 40 kHz tone, which the sensor picks up when the signal is reflected back from the object. The time between the transmission and reception of the sensor is noted, and the distance is determined using the formula. The distance was calculated as (Time to Receive Reflected Signal ${\times}$ Sound Speed (340 M/S)) / 2. The formula is then used to convert the distance to centimeters. Speed of Sound (cm/s) ${\times}$ Time Duration (s/2) equals the distance in centimeters. The drainage water level can be calculated by deducting the distance from the drainage height. Fig. 7 shows the observed values of various drainage parameters reported by various sensor nodes. Each container houses four drainage parameters: toxic gas level indicator, air pressure indicator, drainage water level indicator, and drainage access chamber status indicator. There are four containers for four sensor nodes. The water level indicator readings are represented in red when the drainage level exceeds 50\% of the drainage height.

Fig. 7. Result displayed in the NCC simulation panel.

5. Conclusion

Implementing a 3D model for designing and optimizing drainage systems in smart cities offers valuable insights into system performance. The model saves time and money while ensuring efficient drainage by identifying potential issues before construction. Integrating sensors, such as the MQ-4 gas and HX710B atmospheric pressure sensors, enable real-time monitoring and data collection. Future prospects include IoT integration, automated control systems, climate resilience, smart urban planning, public engagement, green infrastructure, and water recycling. These advances can produce sustainable, resilient, and efficient smart drainage systems for improved living standards in smart cities.