Description:AI BASED SMART BIN FOR MUNICIPAL SOLID WASTE SEGREGATION

PROBLEM IDENTIFICATION
Municipal solid waste management is the huge problem faced by government. As the population increases the amount of waste generated by the people also increases. Collected waste of any form mostly consists of non-biodegradable waste and bio-degradable waste (some at decomposed form). Since, waste is present in a combined form, separation needs huge amount of labour than any other process and requires quite a lot of time.
• Under the Waste Regulations 2011, waste must be segregated paper, cardboard, plastic, metal and glass at source unless it is technically or economically unfeasible.
• Under the same regulations, it has been implemented the waste hierarchy; reduce, reuse, recycle, other recovery and disposal. By law, it should be established this hierarchy and segregation helps with recycling in particular.
• India’s waste management sector is expected to be worth US $13.62 billion with an annual growth rate of 7.17 percent. Much of the waste India produces simply ends up in landfills without proper processing or treatment - redirecting this untapped waste to proper treatment and processing facilities will open up new investment possibilities.
According to survey, in only six years 9,730 people have died from multiple chronic conditions emanating from their employment in the dehumanising practice of manual insulation of trash segregation in India. Such workers are prone to confront a few health issues due to their working environment. The physiological health issues include injuries, gastrointestinal disorders, musculoskeletal disorders, respiratory disorders, dermatological disorders, ophthalmic disorders and infections such as gastroenteritis, E. coli diarrhoea, cryptosporidiosis, giardiasis, hepatitis A, hepatitis B, leptospirosis, salmonellosis, typhoid fever and tetanus.
Segregation is the primary activity that needs prime importance to enforce effective waste management in urban areas and to reduce the size of landfills gradually and eventually land in a “no landfill” scenario in four to five decades from now. Even though many inventions related with waste processing were emerged, they were only expected to be a disposal system. Waste treatment plants which consume enormous transportation and treatment expenses are not that effective on the segregation. The waste can be utilized as a best resource with the help of such proposed method.

EXISTING METHODS
In general, municipal solid waste management is done by the following ways
Incineration: Combustible waste from households and waste wood that is not suitable for recycling undergo thermal treatment in waste incineration plants or waste wood furnaces. The heat released in the process is used to generate electricity and heat buildings. Waste with a high calorific value and low level of pollutant contamination can be used in industrial plants, e.g., cement plants, as an alternative to fossil fuels. Waste that is contaminated with organic pollutants undergoes separate thermal treatment (e.g., in hazardous waste incineration plants). Incinerators must have a flue gas treatment system. The requirements for flue gas treatment and the incineration system are based on the nature of the waste. Specialised waste disposal companies treat the waste in accordance with the requirements of the incineration plant. This guarantees that the fuel will be of a high quality and reduces the accident risk. The companies ensure, for example, that no undesirable reactions occur when liquids are mixed. Waste materials that are used as substitute fuels in cement plants must be crushed in advance and set at a constant calorific value.

Chemical-physical and biological treatment: The objective of both chemical-physical and biological treatment is to enable the removal of pollutants from waste or its safe landfilling. Wastewater and polluted excavated material are typical of the types of waste that are managed in this way. Following chemical-physical treatment, the pollutants can be disposed of in concentrated form in facilities suitable for this purpose.

Landfills: Residues from waste incineration or waste that is not suitable for material recycling or thermal treatment are deposited in landfills that are compliant with the legal requirements. If the waste does not fulfil the requirements for landfilling, it must be pre-treated.

Collection and logistics: The waste management sector involves many different specialised actors. Their tasks include the collection of waste at source (industry, commerce and households) in suitable transport containers, its intermediate storage and handover to waste disposal operations. The treatment of waste is often based on a cascade of specialised plants. In all cases, smooth logistics are a precondition for the efficient management of waste. In the case of hazardous waste, in accordance with the Ordinance on Movements of Waste, the handover must be documented.

PROPOSED METHOD
The solutions that considered were of diverse nature. Initially, the water floating method was considered. Here, the light weight of the waste materials contributed to the separation but was not effective enough as the weight of the material became a main factor of consideration. Further exploration led to the thought of sensors. Later was it discovered that there were no such sensors that could differentiate waste materials taking into account its texture, size or external features.
Finally, the solution arrived at is a combination of various ideas that was proposed, providing a holistic approach towards the main goal of waste material segregation. Firstly, the waste is dumped at the opening of the bin (10) which houses the counter rotating crusher setup (5) with appropriate amount spacing between them. Then the trash is passed through the rotating tunnel (12), where mild heating and rotating takes place in order to separate the coalesced materials. The segregated waste is the made to fall on the conveyor (4) arranged below this tunnel. The camera module (16) which is focused at the passage of conveyor. It is trained in such a way that it detects the sanitary wastes from this municipal waste. The servo mechanism (17) is triggered when the sanitary wastes are detected and direct them into one of the waste collection bins (1.1).
An air blower setup ((14) and (15)) is used to remove the light weight paper waste from the conveyor to another container (1.2). At the end of the conveyor a permanent magnet (22) is placed to attract the magnetic materials and remove them into the separate container placed below it. Residual wastes are removed into different container. The weight and level of the bins are identified with the help of the IR-proximity sensor and load sensors. These values are sent to the online waste management website with the help of GSM module. The location of each bin is updated with the help of GPS module.
The Solid Waste Segregator is a system designed to automate the process of waste segregation. It uses advanced technology and sensors to separate waste into four distinct categories:
• Sanitary waste,
• Lightweight waste (papers and wrappers),
• Magnetic waste, and
• Non-magnetic waste

The proposed method provides the following advantages
• Compact system which can replace the conventional municipal green dustbins
• Intervention of Manual labour for segregation of waste will be totally avoided
• Eco friendly solution for both man and nature will be provided
• Requires less energy for segregation
• Collection and management process can be improved
• Revenue can be generated from the segregated waste

BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 illustrates the block diagram of the system
Figure 2 illustrates the isometric view of AI based smart bin for municipal solid waste segregation
Figure 3 depicts the entire conveyor arrangement with blower, servo mechanism, camera module and permanent magnets
Figure 4 reveals the rotating tunnel setup
Figure 5 reveals the opening of the AI based smart bin for municipal solid waste segregation
Figure 6 illustrates the crusher setup
Figure 7 illustrates the circuit box with bins, IR-proximity sensor and load cells
Figure 8 illustrates the 2D drawing of the AI based smart bin for municipal solid waste segregation with dimensions
Figure 9 depicts the prototype of the AI based smart bin for municipal solid waste segregation
Figure 10 illustrate the interface of online management portal

The entire waste segregation setup is the compact one which can be identified in Figure 8. Each process involved in the system were explained in detail below
Waste Identification and Crushing:
The processing begins with loading the waste into the opening (10) of the bin, wherein a pair of counter rotating crushers (5) are available with appropriate distance between them to tear up the tied packages of municipal waste. This crusher consists of various forks focused in different directions which basically leads to tearing and breaking the solid wastes present in the mixture which can be seen in Figure 6. Each crusher roller (5) is driven by a separate AC induction motor (6) to provide high RPM and torque for crushing. Both crushers have a timing pulley (7) which connect the motor (6) via timing belt ((8) or (9)) individually. Simultaneously the waste identification is done by IR-proximity sensor (20) which placed at the top of the bin opening (10) to turn on the entire setup as shown in Figure 5. The bin will remain ideal until the waste is identified at the entrance of the bin.

Rolling Tunnel:
The waste is then directed into the rotating tunnel (12) which is supported on the frame with the help of slider (11) held below the crusher setup (5). This rotating tunnel (12) is inclined at 15 angle to provide both centrifugal and gravitational force for the waste to provide proper segregation into individual components. It also encompasses the mesh arrangement as shown in Figure 4 to remove tiny dust materials separately. This rotating tunnel is driven by the DC Gear motor (13).
Simultaneously, a mild heat is applied on the rotating tunnel for drying up the moisture content present in the waste material with the help of blower. This phase is indicated as a ‘slow process’ as the whole processing is done with proper care with regards to segregation. After the whole waste processing in the tunnel, the waste is directed towards the Conveyor Belt (4).

Conveyor and segregation:
The separated wastes are set to fall into the conveyor belt (4) with the help of gravity. This conveyor is driven by a DC Gear motor (13). As shown in Figure 3 the conveyor is supported with rollers at both ends. A camera module (16) is placed at the starting end of the conveyor to detect the sanitary waste materials like sanitary pads, masks, syringes of domestic use. This was achieved by training the raspberry pi microprocessor which is placed inside the circuit box (18) with OpenCV and machine learning algorithms. The images used for training as a data set were exclusively prepared by the inventors.
This camera module (16) is connected with the servo mechanism (17) via raspberry pi. Whenever the camera detects the sanitary waste, a trigger signal is sent to the servo mechanism to segregate into one of the four bins (1.1) below the conveyor level. This bin arrangement can be seen in Figure 7.
Further the wastes are carried through the conveyor (4). A blower (14) placed next to the camera module (16) is made in such a way that it blows the light weight wastes (papers and wrappers) with the help of blower pipe (15). Such blown waste is directly dumped into another container (1.2) placed opposite to the blower pipe.
Magnetic waste materials which present in the remaining mixture are carried through the conveyor as an effect of magnetic attraction with the permanent magnet (22) placed between the conveyor layers and dumped at the bin (1.3) where the magnetic attraction gets week. Residual wastes which contain wood, plastic bottles, stones and clothing items, etc., move till the end of the conveyor. Waste materials falls directly into this container (1.4) due to the effect of gravity, thus these general solid wastes form municipal wastes are collected separately.

Autonomous control:
Every container is fitted with load cell (2) and IR-proximity sensors (21) to indicate the amount of waste and space available in container. These readings of load cell and IR readings were used for double checking of the level of each bin. The load cells (2) are connected with the wooden board (3) on which the waste containers will be placed, instead of connecting directly into the bins. This arrangement makes the collection of waste effortless.
The data from each load cell (2) and IR-proximity sensor (3) will fed into the raspberry pi module which is trained to identify the level of the bins with the inputs like volume of the empty bin, level and weight readings after the dumping the waste into the bin. Finally, the level information of all bins was sent to online waste management platform as shown in Figure 10 to the respective management personal with the help of GSM module connected with the microprocessor. The online platform also contains the location information of each segregation system which can be identified by the GPS module connected with system. The alert message also sent to the respective management persons upon the filling of the bins.

Collection of wastes:
Placing this segregation system as a replacement in the places of conventional green dust bin the waste segregation can be done with low segregation and transportation charges, as the segregation happens in the source itself. Also, the detachable bins of (1.1, 1.2, 1.3 and 1.4) provides the conventional option for collection.
The segregated waste of different forms like magnetic, sanitary and lightweight wastes were sent directly to recycling process. Another segregated category of waste which can be sent to further processing of any one of the above-mentioned existing methods.

Channels:
? Government Waste Transfer Station, all over India.
? Private Waste Transfer Station all over India
? Tamil Nadu Pollution Control Board
? Manufacturing Industries
? Agro - Industries

Compact design:
The proposed idea is a compact design due to which it can be placed at different public places such as
? Malls
? Government offices
? Apartments
? Colleges
? Private colonies

Unique advantages:
? The idea suggested here helps us to reduce the amount of segregation time, makes the work a lot easier and can be used in small and large scale as well.
? The project will help mankind to segregate trash in automotive system as it makes it easy to segregate a range of elements.
? This project paves the way to the concept of reuse, reduce, repurpose, recycle and thus, converting waste into products of agriculture utility.
? These recycled products are found useful in other domains too.

Key metrics:
? Reduce the land occupancy
? Protects the environment
? Turning waste to treasure
? Saves valuable time and resources
? Cleans the landscape
? Reduces pest infestation
? Adds to waste energy
? Returns from Investment

Revenue stream:
This compact design product is an effective solution and also promotes the idea and supports the idea of Swatch Bharat mission. Clear awareness can be provided on community-based waste management programs and environmental management programs. After implementation, this will remain to be an effective solution and will be of great help to sanitary workers. , Claims:We claim:
Claim 1: An AI based smart bin for municipal solid waste segregation, which comprises crushers (5), a proximity sensor, a rotating tunnel (12), a conveyor belt (4), a camera module (16), servo mechanism (17), air blower setup (14), permanent magnet (22) multiple storage containers (1); each connected with volume identification and weight sensors, a microprocessor to perform entire process, and communication devices to connect them with online management portal.
Claim 2: The crushers (5) of the segregation system as mentioned in claim 1 have a pair of counter rotating mechanism which have individual motors to drive them.
Claim 3: The rotating tunnel (12) which mentioned in claim 1 is driven by a DC Gear motor and has a mesh arrangement on it.
Claim 4: The microprocessor which mentioned in the claim 1 is the raspberry pi 4.
Claim 5: The camera module (16) which mentioned in the claim 1 is trained with OpenCV and ML (machine learning) algorithms to detect sanitary wastes by the given data sets, and connected with raspberry pi microprocessor to control servo mechanism (17).
Claim 6: The level and weight sensors which mentioned in claim 1 are IR-proximity sensor (2) and load cell (21) respectively and they are connected with raspberry pi microprocessor which mentioned in claim 4.
Claim 7: Proximity sensor which mentioned in claim 1 is an IR-proximity sensor (20) and connected with raspberry pi which is in claim 4.
Claim 8: The communication devices which mentioned in claim 1 are GPS and GSM modules and connected with raspberry pi which is mentioned in claim 4.