AIoT-Based Smart Bin for Real-Time Monitoring and Management of Solid Waste

In the current time, the immense growth in population creates unhygienic environment for the citizen of a society with respect to waste generation. +is rapid generation of waste leads to various infectious diseases in the environment. As followed by the traditional municipal system, in our surroundings, we can see over flooding of solid waste in the garbage bins. Solid waste management is a pivotal aspect in traditional systems and it is becoming dangerous in most populated areas. Arduous labor works and costs are required to manage and monitor garbage bins in real time. To maintain the cleanliness of a city and for real-time monitoring of trash bins, a smart bin mechanism (SBM) for smart cities is proposed in this paper, which is based on Artificial Intelligent of+ings (AIoT).+e SBMworks on the 3R concept, that is, Reduce, Recycle, and Reuse.+e SBM has the access to get real-time information about each bin and avoid overloading of these bins. +e proposed framework reduces the labor cost and saves time and energy of the system. It also reduces the rate of disease infections by keeping the cities clean. Fuzzy logic is used for decision-making in selecting appropriate locations in the cities to install trash bins. +e framework is implemented in the multiagent modeling environment, that is, NetLogo.


Introduction
e significant growth of the Internet is leading to the emergence of new technology, that is, Internet of ings (IoT) [1]. e term IoT was first used by Ashton in 1999 [2]. IoT is the hub of physical devices that are interlinked through the Internet. ese physical devices, that is, sensors, RFID tags, and various intelligent nodes, can communicate at any time from anywhere. IoT is the backbone of future communication systems where everything will communicate and share information smartly without human instructions [3]. e interconnected devices are transformed as smart objects, which have computational skills that are used to monitor an environment leading to smart cities.
IoT promotes various application areas, such as smart health, [4][5][6][7][8][9][10][11][12][13] smart city, [14,15], environment monitoring [16,17] smart home, [18,19], traffic management [20][21][22][23][24], smart education system, [25], smart farming, [26,27], and many others [28][29][30][31][32][33][34][35]. In smart cities, various problems occur when devices communicate with each other; one of the important problems is waste management. e main causes for this problem are the rapid growth of the urban population, high demands for food, and various other factors that are influencing the environment in smart cities. With an increase in population, the management of waste or garbage is a very hectic job to do in the current time. Being a member of the society, every local house, industry, and factory generate some amount of waste on a daily basis. is waste is ultimately collected in waste bins and eventually collected by the municipal vehicles and moved to dumping areas for disposing or recycling processes.
To keep the environment green and clean, monitoring and disposing of waste is very important these days. Improper disposal and poor monitoring of collected waste and waste bins can cause serious damage to human lives.
is waste can spread various life-threatening diseases that in turn harm the lives of a whole city and country as well.
Nowadays, cities are facing various problems, such as small parking spaces [36], waste management, communication barriers in traditional systems, and health issues to name a few [37,38]. All these problems directly affect the living of humans in their daily routine lives [39]. To overcome and solve the existing problems, a new concept has emerged in the light of IoT, named smart city [14]. IoT provides various new services in a smart city and develops an intelligent society [40,41]. In IoT-based smart cities, physical devices interact and provide ease to humans according to their own intelligence [42]. IoT is further divided into various fields presented in Figure 1 [43], which explains the web-based rankings of each area in percentage. Based on the literature study, smart homes and smart cities get the highest rankings, which shows the trends in the modern era of technologies. e rapid growth in population and generation of daily routine garbage or waste make the environment unhygienic for the citizens. e waste is divided into two types, that is, wet and solid waste [44]. In this paper, the focus is on solid waste management. erefore, a waste management mechanism is proposed for smart cities, named SBM (smart bin mechanism), in order to sanitize and clean the environment intelligently. It is designed for solid waste management and recycling of waste because waste is recyclable and can be reused. e waste management procedure comprises five steps that include collecting waste, transporting, analyzing and processing, recycling, and disposing [44]. In SBM, smart bins (SB) are installed in the urban areas at different points that store garbage. Primarily, the SBM is designed for real-time monitoring of the garbage collecting points.
e proposed system will reduce the labor work, time, and cost that are very high in the traditional garbage collection system.
In SBM, different entities have been used which show their own roles in providing services to the citizens. It has trash bins, trash collecting vehicles, and a central database to keep records of the levels. All the entities and their roles are discussed in Section 3.

SBM Contributions.
Waste management is an emerging era of most populated as well as less populated cities. In IoT, it also has emerged as a field in smart cities. Various prior research has been conducted for collecting waste, but most of the research is server-based or authority monitoring of garbage bins that are installed in public places. ese systems act upon receiving requests from garbage bins and sending collecting vehicles toward the requesting bins. is mechanism consumes enough an amount of energy and time on fulfilling of a single request. erefore, an intelligent edgenode based mechanism is necessary for collecting waste from requesting bins, which consumes less energy and time.
e proposed framework is based on edge-nodes, that is, the trash bin. When a trash bin reaches the threshold level, it makes a request directly to the collecting vehicle instead of forwarding the request to any central authority. e proposed framework has a significant role in waste collecting procedures as it consumes low power because of its novel edge-based mechanism. Time consumption is also low in SBM because of the less iteration from the request-to-response procedure. Some significant contributions of the proposed framework are as follows: (i) Proposing a smart bin mechanism that is based on IoT technology and applications (ii) Real-time monitoring of the trash bins in a smart city (iii) Using trash bins in an effective way to facilitate the municipal department and citizens as well (iv) Reducing labor cost and optimizing resources (v) Improving environment goals and cleaning cities with limited resources e proposed study is novel in terms of real-time monitoring and decision-making using fuzzy logic processing. Fuzzy logic provides the best suitable and less dense site of the city to install trash bins. Two fuzzy parameters are used in fuzzy inference systems, that is, distance from the collecting/dumping zone and access to the trash bins. e proposed study is beneficial in the future era of modern technologies, where everything will be connected via the Internet and communicate without human interruption.
e study supports the smart city concept by providing real-time monitoring for climate change. It provides a decision-making mechanism by using a fuzzy inference system. e rest of the paper is divided into five sections; that is, literature study is discussed in Section 2, SBM methodology is presented in Section 3, results and discussion are provided in Section 4, the paper is summarized in Section 5, and challenges and future work are elaborated in Section 6.

Literature Study
For the last few years, many researchers are focusing on IoTbased applications, especially smart city [45]. According to [46], a smart city is an infrastructure where everything is interconnected and can interact with each other. In a smart city, everything is supposed to be smart and intelligent in decision-making ability [47]. A smart city leads to a smart environment [30], smart health [48], smart parking, smart economy, smart administration [49], and smart living of the people [50]. e smart city provides all the better facilities to citizens and assures that there is a clean and green environment for them [51]. To make the environment clean, there should be an effective system for collecting waste. In this section, various research about garbage or waste collection and a better management mechanism for the collected waste is reviewed.

SWMS.
Waste management systems play a vital role in reducing the unhygienic objects from a particular area. To avoid these conditions, a smart waste management system (SWMS) [44] is proposed that is based on IoT technology. e SWMS consists of public garbage collectors with embedded technology that is used to monitor real-time level of garbage bins in public places. Based on the level of garbage bins, an optimized path is selected for the garbage collecting van that eventually reduces fuel cost. e SWMS divides garbage bins into master and slave bins. Each garbage bin is composed of three sensors, such as level sensor, humidity sensor, and load sensor. e master bin is continuously transmitting its data to the cloud with the help of Wi-Fi. rough real-time monitoring, accurate reports can be generated, and, therefore, the efficiency of the system is enhanced.

GMS.
Overflowing of dust bins at public places increases unhygienic environment for the people, especially, in developing countries; this creates serious health problems for the citizens. To cope with these types of situations, an IoT-based garbage monitoring system (GMS) is proposed in [52]. e system contains various dust bins that are distributed in the city. Dust bins transmit the data to concerned authorities in order to clean the garbage. e block diagram of the proposed model consists of two sections, that is, transmitter and receiver sections. e transmitter is installed in dustbins, which is used to transmit collected data from sensors to the receiver end. At the receiver end, the central system receives the data sent from the dustbin and processes it accordingly. e authors used Raspberry Pi, RF receiver, and a web browser to fulfil the requirements of the system. e proposed system has some limitations in terms of lacking in reliability of communications among different modules.

IoT-Based SWM.
Waste management is an important service provided by smart cities and supported by IoT. An enhanced system for waste management is proposed in [53] by considering the growth of the population in urban areas. e proposed model mainly consists of four entities, such as smart bins, waste areas, management centers, and collecting trucks, as shown in Figure 2. Statistical analysis and decision-making are successfully done based on the data provided by the above-mentioned entities. e authors stated that the proposed model overcomes the existing issues in the waste collection process, that is, location issues, cleaning costs, health hazards, and many others related to waste management.

SWM by K-Query
Scheduling. An IoT-based system is proposed in [54] that is used for waste management with the help of K-Query scheduling.
e system is composed of microcontroller module, GPS module, and ultrasonic sensor. ese modules are installed in trash cans. e sensors are used to monitor the trash cans. When a trash can reach an appropriate level, the sensor calculates the level and transmits it to the cloud through Internet. e K-Query scheduling is used to store threshold values in a table created in the MySQL database. e architecture of the system is shown in Figure 3. e K-Query is helpful to reduce unknown entries in the database. For a shorter path, a code with a map and location point is executed only one time. ere is no need to execute the code in order to find the route for every event. is system is helpful in reducing manpower used in collecting waste from different locations using manual systems. However, the system has lacking with respect to power interruption.

SWC as a
Service. An IoT-enabled solid waste management system is proposed in [55] for monitoring garbage bin and dynamic routing of the garbage collectors. e proposed system consists of an embedded device for real-  time monitoring and scheduling of routes for garbage trucks. e architecture of the proposed system is presented in Figure 4. A mobile application is also designed for the truck driver to handle the data coming from garbage bins and further transmitting it to the cloud. In the proposed system, two garbage bins are installed in one place and solid waste is segregated from dry and wet garbage bins. e system is efficient in decision-making as it uses the GPS module and Google Map API for finding the optimal route to reach the garbage bins.

SCGCMS.
In SCGCMS [56], a smart waste management and monitoring system is proposed for public waste collection that is based on IoT technology. e system consists of two phases where, in phase one, dustbins are installed in different locations and filled randomly while, in phase two, the route for collecting vans is decided on the basis of the dustbin filling ratio. e filling threshold is defined by the authors, which is 10 cm per dustbin. e system architecture of SCGCMS is shown in Figure 5. In this system, genetic algorithms are used for gathering waste. e dustbins are composed of a weight sensor and Raspberry pi Uno board that is connected with GSM modem and ultrasonic sensor for communications.

ML-Based WMS.
With the rapid growth of IoT and its applications, various critical issues have appeared in todays' lifestyle. One of the most critical issues is waste management in urban areas. To reduce these types of issues, a waste management system [57] is designed for a campus of the Ton Duc ang University in Vietnam, which is based on machine learning (ML) in the IoT environment. e authors used graph theory and ML that provide optimal path selection for waste collection on predicting the probability of garbage in trash bins. e proposed system is used for realtime monitoring through integrating multichoices, that is, ultrasound distance, E32 TTL-100 433 MHz with the LoRa spread-spectrum technology. Energy supply to each node in the network is provided by different sources, that is, solar and batteries. e proposed system for waste collection is better than the existing systems in terms of optimal pathfinding and flexibility. Table 1 presents the advantages and limitations of the surveyed schemes.

Proposed Methodology
In light of IoT technology, waste management is an important service that is supported by IoT. In today's time, waste management is a collective issue in most countries, which needs uninterrupted importance for management. In traditional waste management systems, the rapid growth of garbage leaves the public places unhygienic and dirty. e unhygienic environment can cause various deadly diseases. e prior research focused on the centralized system for waste management that is managed by a central authority. In this study, we are proposing a smart waste management system for real-time monitoring of "trash bins" in order to take timely actions for cleaning the bins and maintaining a disease-free environment for the people.
e proposed system is based on edge-nodes, that is, trash bins. In this  system, a trash bin is working as an intelligent node in the entire processing of waste management. e smart bin mechanism is illustrated in the form of a block diagram in Figure 6.  When the level of the bin increased to level >� 90%, its color turns into red, which is the sign of a full bin.

Trash Collecting
Vehicle. e vehicles that collect waste from the trash bins are dependent on the populated areas of smart cities. Mostly, smart cities are overpopulated which leads to difficulty in the collection of waste from densely populated areas using the same size of collecting vehicles. Each TCV is connected with the central database from where it gets information about the requesting TB. e TCV collects waste from the trash bins and brings it to the dumping zones for further treatment.

Central Database.
e central database is used as an information center as well as a storage point, which contains each and every detail of the TBs, TB-IDs, TCVs, their locations, and every single detail about these entities. Whenever an event occurs in the system, the processing information and status of trash bins are stored in the database.

Transmission Pattern of the Proposed System.
e whole mechanism of collecting waste is described in Figure 7, adopted from [56]. When a trash bin gets filled or reaches its threshold limit, it changes its color into red and transmits a notification to the TCV through a gateway. e TB notification consists of TBL, TBC, and TBW. e TCV receives the request from the TB and forwards the status of the requesting TB to the database for updating. e TCV collects waste from the filled bin for further treatment, such as disposing waste or recycling and reusing waste. e TCV updates the status of requesting TB in the central database after collecting waste from the bin. e whole processing of the system is supported by Reduce, Reuse, and Recycle mechanisms.

Processing of the Trash Bin.
e trash bin checks the level of waste. If the level >� 90%, the TB changes its color into red and forwards a request to TCV for the cleaning process. If the level <90%, the TB color remains green, and without forwarding request, it rechecks its level and so on.
Generally, the proposed system is a repetitive mechanism that consists of the following steps: collecting waste, planning and analysis, segregating waste at the waste plant, and recycling or disposing of waste. e hardware structure of SBM consists of TBs, which are installed at different locations of the city having their unique IDs. At the initial level, each TB is green in color, while the weight and level of TBs are recorded accordingly. Once a TB reaches its threshold level, the color of that particular TB will change into red with obvious measurements of weight and level of waste in the percentage form. e TCV is another significant

Scheme
Contribution Limitation SWMS [44] Path optimization for garbage collecting van Failure of sensors leads to system failure GMS [52] Fast transmitting mechanism for garbage collection Lacking in the reliability of communications IoT-based SWM [53] Provision of statistical analysis Lazy transportation affects all four entities SWM by K-Query scheduling [54] K-Query scheduling is used for database management False monitoring can be a disadvantage for K-Query scheduling SWC as a service [55] Utilization of mobile app to facilitate van drivers Bandwidth constraints of cloud can affect the mechanism SCGCMS [56] Uses genetic algorithms for collecting waste Scalability issues can occur with complexity of the system ML-based WMS [57] Flexibility and optimal path selection using machine learning Failure of batteries or interruption in solar provision  6 Scientific Programming factor in the proposed system. e TCV is informed by the filled TB to collect waste from it and updates the database with the new status of the requesting TB. e CDB is the central database in the SBM model. e duty of CDB is to store data regarding the location of the TB and TCV with the optimal route for waste collection. In SBM, for each event that occurred in the processing of waste collection and management, data for each process is also updated on the cloud. Eventually, the cloud contains information about each single event of the entities that are participating in the system. Working of the SBM is described in the form of flow diagram, as shown in Figure 8. e main function of the proposed system is "trash to cash" that is based on three concepts, that is, Reduce, Recycle, and Reuse. "Reduce" refers to minimizing the amount of waste in the smart city as a particular situation of SBM. "Recycle" refers to recovering or reprocessing of dumped material that is extracted from the trash bins. "Reuse" refers to utilizing the dumped material after its recycling process. e processing of the system is started when an event is created in the trash bin. A threshold limit is set for each TB that helps easily accessing the bin and starts the cleaning process for quick and fast service provisioning.

Trash Bin Control Using Fuzzy Logic Processing.
e fuzzy set theory plays a significant role in real-time scenarios to make decisions. Fuzzy logic was first introduced by Zadeh in 1965 [58]. e fuzzy logic is beneficial to deal with vagueness and uncertainty in real-time monitoring of the environment. e fuzzy expert system (FES) is composed of three fundamental steps: fuzzification, inference rules, and defuzzification. A fuzzy expert system is the combined form of membership functions, if-then rules, and fuzzy operators. Moreover, FES is a mapping between the input and output values. Fuzzification converts crisp input values into fuzzy input, fuzzy rule base and/or knowledge base apply appropriate if-then rules, and defuzzification reconverts the fuzzy output into crisp output or in human readable form. e basic architecture of FES is presented in Figure 9.
In this study, fuzzy logic is utilized for decision-making in selecting appropriate locations to install trash bins. Fuzzy logic provides a better reading of the trash bin in real-time monitoring of the situation by using different levels of linguistic values. e linguistic variables have three levels of readings with different categories. We used three attributes for acquiring data of each trash bin, that is, TBL, TBC, and TBW, which indicate the trash bin level, color, and weight, respectively. Table 2 describes the input values of the trash bin level with categories low, medium, and high.
e input values for the attribute trash bin color are categorized into red, yellow, and green, and their ranges are defined in Table 3. Table 4 shows the ranges of input values for trash bin weight with categories of light, medium, and heavy.
One attribute, that is, trash bin status (TBS), is used for the output readings of the trash bins. e TBS is classified into three stages, that is, bad, average, and good. e output ratings are described in Table 5. Fuzzy logic helps in decision-making in such types of scenarios while selecting appropriate locations and size for installing trash bins. In this scenario, two attributes, that is, access to heavy vehicles and distance from collecting points, are used for choosing suitable places to install trash bins. Based on these two attributes, the system makes decisions and these are divided into different categories described in Table 6.   ese parameters are helpful in selecting the best suitable site in the city to install trash bins for the citizens. e proposed system is reliable in terms of providing consistent services to the municipal department for monitoring and cleaning the city in real-time scenarios. For decision-making, the discussed attributes are used for generating fuzzy rules, such as if-then rules. ese rules are then integrated to form an output value, which is demonstrated in Figure 10. Two input variables are used to choose a suitable location for the TB installation, which gives a single output by applying inference rules.

Results and Discussion
We have performed simulations of the proposed framework in real-life experimental environment with different test runs of loading and unloading of trash bins. e NetLogo platform is used to implement the smart waste management mechanism by using different simulation cases in time T (minutes). At the initial stage, 20 to 25 trash bins are randomly distributed in a smart city with TBL � 0% and TBC � green. On tick 1, TBL � 10% and TBC remains green. When TBL � 90%, TBC turns into red that is an alarming situation from the bin to a vehicle and thus the bin makes a request to the nearby collecting vehicle for the unloading process. For tick 1, the simulation time T � 0. In the first case, 10 trash bins are distributed in time T � 45 minutes. Figure  11 shows the distribution of bins where the reading of each bin is recorded with different waste levels as per tick. e waste level in each trash bin is recorded against the total number of bins in the experimental environment. Each TB has some attributes, which are integrated to find the status/ readings of each bin. e process is shown in the following equation: For a test run, one TB is selected as a testing bin and with waste level or level of filling, different rounds are performed on that particular bin. In Figure 12, 8 different rounds for a TB are shown that are directly corresponding to the level of filling the bin. ese records show different readings on throwing action of garbage in a bin. e binunload operation is performed when a bin is full to its level. In Figure 12, R denotes the readings of the selected testing bin. e measuring weight level of a testing bin recorded with respect to time is described in Figure 13. It shows the weight of solid waste in a bin with regard to kg per time T. It shows the readings of throwing waste as per operation or interval. e weight of waste/garbage that is thrown in the TB is measured in kilograms against the measured time T in which a citizen throws it in the TB.            Two fuzzy parameters are used to decide an appropriate location for installing trash bins in the city. Fuzzy parameters are helpful in selecting the size of the trash bins depending on the density of the area, such as less, moderate, and high dense areas.

Conclusion
Conventional solid waste management systems have several shortcomings in terms of late unloading, hindrance in new techniques, lacking in throughput, less access to actual data, and many more. erefore, an advanced approach is the need for the time to overcome all existing problems in the waste collecting process. Generally, waste collection has more consumption of cost from the municipality budget. In this paper, a real-time smart bin monitoring framework is proposed to get real-time access to data from the bins and implement the collecting procedure accordingly. e proposed framework is achieved by using a theoretical and architectural model. e model is implemented in a real-time environment of NetLogo and the experimental results show that the proposed framework is very responsive and effective for the environment. It is also effective for the economical aspects as it reduces the cost of labors and fuel cost of collecting vehicles by minimizing their extra visits in checking bins' status. Once a bin reaches its threshold limit, it informs the collecting vehicle for the cleaning process, which saves time, cost, and energy. e SBM is userfriendly as it obstructs the overflow of bins. It is useful for IoTbased smart cities, which helps to keep the environment clean and disease-free for the citizens. e SBM is supportive for realtime scenarios by using fuzzy logic processing in order to designate trash bins according to space and density of the environment in public areas. Fuzzy logic helps the system in selecting the best fitted site for each trash bin. Generally, fuzzy logic boosts the system for performing effectively in the environment.

Challenges and Future Work
ough the SBM framework designed for smart cities in the context of IoT has potentials, at the same time, it has the following challenges: (i) Distribution of trash bins in the most populated areas where the amount of waste is unpredictable on daily basis (ii) Disturbance in the Internet connectivity due to various causes, that is, weather disruption or defected connection (iii) Lazy transportation: traffic jam could be a big challenge for vehicles to reach on time and collect garbage (iv) Communication between two entities and damage of batteries could be severe challenges for the system In the future, the model may be extended to an alternate and the shortest pathfinding for collecting vehicles in order to enhance transportation and remove collecting barriers. In addition, adding alternate sources for connectivity in case of power failure or weather hindrance may also be considered. Further, to facilitate the mechanism and save more energy, automated segregating TBs can be installed for dry, wet, and hazardous types of waste.

Data Availability
is research is based on simulations, which are performed in a simulator. erefore, there is no dataset used in this research.