Internet and its growing applications have become an important part of human lifestyle. As a result of these growing necessities, application of internet went way beyond connecting just the computers to web. One such application is the development of the concept of Internet of things. IoT refers to connection of physical devices to the internet such that they can be automated. The connected things could be anything, such as hardware, software, data or a service. This may include vehicles, buildings, etc. IoT is that development which changed the user-user communication to device-device communication.

Smart city is the city that integrates various technologies and Internet of Things to provide different services of a city [7]. These services may include administration, healthcare, education, transportation, agriculture, etc. The necessity of building smart cities is to enhance the quality of living and the services provided to people. IoT is the need of the future worldwide. It has a wide scope in the near future in every field [9]. One of the fundamental components of smart city is smart environment. Its essential service is the smart waste management. Incorporation of technologies such as IoT transforms the cities into smart cities [13].

Growing waste is a major problem to deal with in today's era. Actions must be taken to develop proper waste management system. In the past years waste management was treated as a static service. Door to door waste collection was a prominent way of waste collection. But with the high technological developments, it is not an efficient approach. Innovative approach is a need for today. IoT provides one such approach in building a real time and automated system. Internet of Things interconnects various objects such that they can exchange data. These things include a wide variety of devices, such as vehicles, buildings, etc. It could be anything like hardware, software, data or a service. The network connectivity it provides makes it more usable. In the near future everything would become smart, i.e. incorporated with IoT. One major concern in waste management is vehicle routing. Directing the vehicles to correct route is necessary. Shortest routes must be chosen always. This will lead to less fuel and time consumption.

This paper provides an innovative approach for waste management. It will help to develop a real time and automated system. City is divided into many sectors. Each sector is incorporated with certain number of smart bins. These bins are included with hardware which consists of microcontroller, GSM, and sensors. Microcontroller is kind of a mini computer that handles the functioning of a system. The sensors will sense the level up to which the bin is filled. Upon completion it will give a notification to the server. The server will route the vehicle to the bin for emptying it. The vehicle will carry the waste to the dumping area. In case of any damage caused to the vehicle, a new vehicle will be routed to that bin. The block diagram of the method proposed in this paper is shown is Figure 1. It shows how every component of the system is connected over the internet. The database is connected to the server as well as android application. This system will also be able to handle vehicle malfunction via the android application.

Figure 1. Block Diagram of Proposed System

Various static as well as dynamic models have been developed for waste management. Although static models cannot handle the dynamic nature of IoT, so most models handle waste collection problem dynamically. This section contains work related to the dynamic approaches.

W. Juyoung et al. [23] proposed a method based on roll on – roll off routing. This approach was suitable for the places that generated huge amount of waste. Such places often could be industrial areas, construction sites, shopping areas, etc. The container size differentiated whether to use roll on or roll off routing. Pickup and transportation were the major concerns in this method. Also unloading the large containers, mainly at construction sites was a problem. Another method was based on Ant Colony System (ACS) algorithm, presented by M. Reed et al. [21]. The ACS is based on probabilistic models and helps achieving dynamic routing. In this approach, the spatial location of bins is considered. Kmean clustering is applied to form partial clusters. It takes into consideration the behaviour of real ants. It also keeps a record of the past actions performed.

K. Buhrkal et al. [5] proposed a method that used routing with time windows which analyze the logistics activity within a city. The time window applies a restriction on the vehicle arrival time which must reach before the window expires. This restriction leads to imposing complexity in the system. This method works towards finding the optimal route for a vehicle such that the waste is emptied into dumping area within the time window provided. X. Bing [4] presented a model which separates the degradable and non-degradable waste. It removes plastic waste from other solid wastes. It is a specialized approach that uses eco-efficiency as an important metric. Eco-efficiency deals with the social issues, costs, environmental impacts, etc. This method considers studying of number of scenario based on various assumptions in collection methods, collection points, vehicle type, etc.

M. Mes [18] proposed a different approach for the waste collection from underground bins. Motion sensors are embedded in the underground bins which transmit information required for dynamic planning. A drawback of this system is that, even though motion sensors are present, the routing is based on static schedules. To achieve dynamic planning, sensor information must be used appropriately. This would lead to a more efficient system in terms of cost, profit and time. Another heuristic solution is devised by Vera Hemmelmayr et al. [12], where the authors state waste collection as a periodic truck routing problem with intermediate two waste depots. Using this system, all kinds if waste, i.e. glass, paper, plastic, etc., are collected and bought to a certain point by the citizens. Therefore, this collection is treated as a node routing problem. This waste is then taken to special sites called Intermediate Facilities (IF). These are similar to depots. Thus, this approach requires a planning period of certain days because every collection point need not be visited every day.

Hui Han and Eva Ponce [11] studied the Waste Collection Vehicle Routing Problem (WCVRP). It studies waste collection of three types of wastes: industrial, commercial, and household. This paper consists of various methods proposed by different authors for waste collection of the three categories. Various classical heuristic and Meta heuristic algorithms are studied. Somayeh Fooladi et al. [6] devised a mathematical model for the reduction of cost of collecting and transporting the waste for recycling process. It keeps a balance and distance between trash cans and the type of waste contained in trash cans to find an optimal route for transport vehicles. Designing of a similarity pattern is done for this purpose and this pattern is used for the selection of route. A mixed-integer non linear programming is applied in this approach. A paper was presented by Aderemi Oluyinka and Olawale Adeleke [1] who gave a new approach for waste management using customer's time windows and road attributes. It works in order to reduce the high cost required to manage the garbage carrying vehicles. Its major contribution was the consideration of attributes of each road in collecting solid waste. Two approaches were proposed, one using nonlinear function and another using a linear function.

P. Pratheep, M.A. Hannan[20] proposed the system for monitoring of waste bins carrying solid waste using RFID technology. It stores all the information related to driver, vehicle, and location in the database. This technology is used for handling the transmission and reception of data. Clients can access the database for system information. The recycle bins are incorporated with reader and the trucks with RFID tag. The hardware implementation of this system is very complex and it is suitable only for designing a prototype due to short range of frequency detection. Theodoros Anagnostopoulos et al. 2] proposed a robust waste collection system for smart cities. They considered two types of trash carrying vehicles, High Capacity Trucks (HCT), and Low Capacity Trucks (LCT). A robust routing algorithm was applied which was common for both the types of vehicles. It could also handle malfunctioning of the trucks. They studied two scenarios for LCT and HCT. First scenario considered carrying the waste from bins directly to dumping are by LCT. Second scenario considered depots as temporary storage, thus carrying waste from bins to depots by LCT and depots to dumping are by HCT. It showed that the use high capacity trucks reduced the operational cost for waste carrying and transportation.

Another method was proposed by S.S. Navghane et al. [19]. The authors presented a model using IR sensors, microcontroller, and Wi-Fi module. This model assures that the bin is emptied as soon as it reaches its maximum capacity. If it is not being emptied then the record is sent to higher authority which can take necessary action. This system also assures that the overall expense required for waste collection is minimized by developing cheaper bins, thus making the system more efficient.

Alexey Medvedev et al. [16] proposed a methodology for smart waste collection. Firstly, they worked for providing Software-as-a-Service (SaaS) product to customers. The customers here are mainly private companies which work for waste collection in cities. Secondly, they worked for developing a system which is beneficial to all the stakeholders involved in the garbage collection process. The important component of their system was cloud based decision support system ( DSS ) , an intercommunication channel. The DSS have all the information related to bin capacity, truck location, fuel available and consumed, etc. This provides an easy way to provide an optimal route to the customers.

Gaurav Kodwani et al. [14] have presented a paper on an automated alarming system using a smart container. Major components of this model were: IR sensors, GSM connected to Arduino, and an alert via SMS. The IR sensors were used to detect the minimum level required. The wireless GSM modem was used for communication over wireless network. SIM card and IMEI number were used for this purpose. Alert notifications were provided through the message on mobile phones.

Vikrant Bhor et al. [3] proposed a smart garbage management system consisting of transmitter and receiver sections. The transmitter section consists of smart bins incorporated with level sensors and microcontrollers. The output of level sensor is provided to the microcontroller. At the receiver end is a control room handled by a person who monitors all the activities. A GSM module is used to send a notification for clearing the dustbin. This system also consists of a GUI which displays the level of the dustbin on a computer screen. This system also assures monitoring of the fake reports, thus making it more reliable.

Meghana K.C. and K.R. Nataraj [17] presented a method for construction of smart bins using infrared sensors and GSM facility. LEDs are present near the bins which emits light when the bin is full. The road maps showing a route to the bins is attached to the internet. On clicking the blinking LED, the road map shows the route to be followed. A text message is sent to the mobile number of the concerned person. All this can be visualized in the control room. This model allows the reduction in the number of times the vehicle must travel to empty the bins. Such smart bins would be a great development toward clean environment.

A solution for waste collection was also presented by Jose M. Gutierrez et al. [10]. The authors developed the intelligent trash cans that can detect the level of the cans filled. The data obtained from these sensors are stored in the database. And this data is used for the further processing for determining routes. The system also preserved the history of the cans, i.e. how frequently the can is filled and emptied. Also routes are provided based on the congestion of traffic. The authors used the shortest path spanning tree algorithm, K- means algorithm, and Genetic algorithms for determining routes. The authors also compared the results between the sectorial approach and dynamic approach. Cost and performance analysis is also done.

Another approach toward the clean environment was presented by Kasliwal Manasi H., and Suryawanshi Smitkumar B. [15]. They presented a system for smart environment via automated waste collection system. Arduiono board, sensors, GSM, and MATLAB GUI was used by the authors. Their system required the master and slave units. Upon filling of the bin completely, a message could be sent to the control room for informing the respective cleaner to empty the bin.

Yann Glouche and Paul Couderc [8] proposed the smart waste management system using the self describing objects. RFID technique was used in this system. This paper aims towards improving the quality if a selective sorting method by gathering information through the RFID tag associated with the waste. The system also separates the organic non-recyclable waste and the recyclable waste. No tags are attached to the organic waste. This information provided by the RFID tags is very useful for the users. The approach deals with developing the smart or digital waste, i.e. those provided with RFID tags. The information can also be stored in the form of QR codes, which can be scanned to retrieve information. But this will incur additional cost in the system.

Another novel approach towards the waste collection problem was proposed by Sruthi K.V., and Manjunath K.N. [22]. They presented a system consisting of ARM LPC2148, along with ultrasonic and pressure sensors. RFID technique was also incorporated to assure authentication. A web page in HTML is also created and an alert system is provided. If the bin is not emptied in time, then an alarm is generated for the user who dumps the garbage into an already filled bin. GSM module was used for message sending and receiving.

The proposed system considers the city to be divided into many numbers of sectors which contains a certain number of homes, public buildings or areas. Within the sectors there are located several bins which serve as the temporary waste storage for that sector. These are called smart bins as they are attached with certain hardware. Each sector has vehicles to carry the trash to dumping areas. A server application is developed which is monitored by an administrator. An android application is available for handling the system remotely. This application is used to track the vehicles, know the status of bins, direct the vehicles to proper routes, etc. Notifications are provided through this app. The GSM technology is used for the mobile communication. Using this a message will be sent to the server, which will also contain a GSM modem. Messages to the mobile number provided are sent via SMS gateway. For every area, a ward member will be appointed to take care that all the activities are performed on time. When a certain level of bin will be reached a message will be sent to the ward member for making sure that the bin is emptied. In case of any delays, this member must take appropriate actions. The flow diagram of the proposed system is as shown in Figure 2.

Figure 2. Work Flow of Robust Waste Management System

Everything is connected over internet. Ultrasonic level sensors are attached to the bins. These sensors in the bin will sense the level up to which it is filled. It will send notifications at different levels of bin being filled to the server, and the server sends this status to the respective person. After the notification of bin being filled to its maximum level, it must be emptied immediately. As shown in figure 2, at 60% of bin filled the message will be sent to ward member, at 85% to the driver and at 95% to the officer. This message will include dustbin ID, area, and the status of bin. As soon as the driver empties the bin, it should update its work done on an android application. This can also be viewed by the officer. Also the drivers can report the vehicle malfunction via the application. This failure reports are also visible at the server application. The admin can then allocate another vehicle for serving the bin.

Microcontroller controls all this functioning. Atmel 89S52, 40 pin dip microcontroller is used here. The RS232 communication is used for communication between hardware and software. The ultrasonic sensor will calculate the level of dustbin filled by the distance travelled by pulse waves. After getting a notification about the bin being filled, the system routes a vehicle to empty the bin. Meanwhile, if any malfunction occurs then the server is informed about it.

The proposed system will also assure that the vehicles used to empty the bins are routed to the smallest route. When a vehicle is serving a bin, it will be informed about the nearest bin to be served, if any. Also the system will handle the truck malfunctioning. In such a situation, the server will be notified about the vehicle being malfunctioned. The server will give all the allotments of this vehicle to another working vehicle. Thus, this system will be able to continue working even under the case of malfunctions.

The hardware to be incorporated in the bins consist of microcontroller AT89S52, ultrasonic sensor, voltage regulator IC 7805, and RS232 serial communication IC, and GSM modem. The voltage regulator IC provides the required 5 V voltage to the microcontroller. All the components are connected over a copper plate. A Printed Circuit Board (PCB) is designed which mechanically supports and electrically connects all the components. A 5 V power supply is needed to be given to the level sensors and microcontroller.

The software section will consist of a server application and an android application. The server application is developed in visual basic dot net. It will consist of an admin login with unique login id and password. Upon login, the server form will open which consist of dustbin number, area, status, message to be sent and the number at which the message must be sent. The status of the bin is also represented graphically on the server side. The server will also be connected to a GSM modem. Messages will be sent automatically upon receiving the bin status via SMS gateway. The server application will also contain an add information form. All the data about drivers ward members, officer and admin can be added and updated using this form.

The android application is developed in Eclipse IDE. This application will be used by officer and drivers. The officer can supervise the system working via this application. He can view the work of drivers on this application. The drivers will use the application for viewing their work to be done. Upon completion of work, drivers can update their work done on this application. Also in case of vehicle malfunction, the drivers can submit the report via this app. It will include driver name, vehicle number, and the area. All this can be viewed on by the officer. Also this information will be updated in the database.

The database is created on a local host. MySQL is used for creating the database. The database consists of tables for dustbin, drivers, ward members, officers, and failure reports. This is connected to the dot net server application as well as the android application.

The hardware to be incorporated in the bins is designed as shown in Figure 3.

Figure 3. Hardware to be Incorporated in Bins

The server application is developed. The screenshots of the server side is shown in figures followed. Figure 4 shows the admin interface. Admin have to login with the login id and password.

Figure 4. Admin Interface

Figure 5 shows the server side after login. It shows the information received from the bins. This includes bin number, area, and the status. Also the message to be sent generates id automatically.

Figure 5. Server Interface

The add information page of the server side is as shown in Figure 6. It includes pages for adding and updating information about driver, officer, ward member, and admin.

Figure 6. Adding Information Page

The graphical representation of the bin status is as shown in Figure 7.

Figure 7. Status of Bin represented Graphically

Moreover, an android application will be developed so as to operate the system remotely. The admin as well as driver can register into this app. All the notifications can be sent via this mobile app. The driver can send the message about vehicle malfunction via this application. Along with the message, the area to which the driver is going will be sent.

The major objectives of the proposed system are to overcome the difficulties appearing in the previously developed methods. This includes, providing of an android application for remote system handling. Also this method includes real time system handling by providing status of the bin at different intervals of time. This justifies the dynamic nature of IoT. Moreover, vehicle malfunction will be handled efficiently. Thus the capabilities of IoT are utilized efficiently.

With the rapid digitization of ever ything, smart environment is the need of today. The door-to-door waste collection is a static waste collection method. This paper presents a dynamic and innovative waste management system via IoT. The smart bins sense the level filled automatically using sensors and sends this information to the server via a GSM modem. The server then sends a message to the respective person. Moreover, the android application will be further developed which will allow remote system handling. Thus, this system will be a reliable and automated system.

The further developments in the presented system can be the development of automated waste segregation bins. This can be done by incorporating a sensor that can differentiate between plastic and non-plastic waste. Also waterproof hardware can be developed to protect the bins from rainy seasons and liquid waste.

This work is a part of the postgraduate level project work and I represent my sincere gratitude to all my teachers for their constant guidance throughout the work and providing excellent atmosphere for Dissertation work.