Every time the garbage is thrown into the dustbin, the garbage will flow to a certain place and must be treated to avoid harm to the environment. The waste treatment process in many cities has been transformed into highly intelligent operation and management activities. Internet of things, ML based waste management platform adds flexible real-time mapping and tracking functions, which can improve waste management results.

Garbage disposal is a great challenge for big cities. Today, the government departments of smart cities such as Singapore, Dubai, Hong Kong, Amsterdam, Stockholm, Tokyo, Melbourne, Seattle, Chicago and Seoul have vigorously promoted the integration of technology into all aspects of their cities. In many cities, the waste treatment process has been transformed into highly intelligent operation and management activities.

Municipal waste management

Today, municipal waste management in any smart city is the interaction of field devices or sensors, which are networked to generate millions of data points. Then, the data obtained from this will be absorbed into the cloud platform, and fed through the complex analysis framework for analysis, and then draw wise and feasible inferences to better serve the residents of the city. The whole process is automated with little human interference.

Waste classification

We all know that waste can be divided into several categories. Paper waste, plastic waste, food waste, water-soluble waste, water-insoluble waste, animal waste, sanitary waste, household waste, industrial waste, etc. Some of them are biodegradable and some are not. Some may even be radioactive waste, which is highly toxic and may cause danger.

Intelligent solutions

We have developed an application that will alarm when conscious citizens capture images of waste on the road or overloaded dumping bins on the road and send them to the cross platform command center. The image received by the command center can then be analyzed using the image point and vector frame analyzer to determine the approximate number and possible categories of different wastes captured in the image. The process does not require human intervention. It uses intelligent algorithms to match past and existing data. Over time, the accuracy of this activity has been close to 90%.

Internet of things driven data analysis and machine learning

The sensor installed on the roadside dustbin can track the garbage collection in the dustbin and automatically send an alarm to the garbage collection truck using the Internet of things integrated system, but the garbage directly thrown on the road escapes the digital vision of the sensor camera. It requires an alert and a conscious citizen or human intervention to cover up the piece of garbage thrown by an abnormal citizen on the road. It should be reminded that alert and conscious citizens must have basic knowledge of photography and be able to skillfully use applications on smartphones.

Data from sensors and images sent by alerts and conscious citizens are completed through complex multipoint and multilayer analysis systems. We use past waste data to train the system on a machine learning (ML) platform to identify and classify waste and roughly evaluate the weight of waste. The ML platform uses waste images collected from more than 60 dustbins in the city in the past at each hour of the day. The machine learning platform is also trained in conventional items usually found in and around dustbins, so they can be easily identified.

Waste disposal capacity management for the last mile

Every time the garbage is thrown into the dustbin, the garbage will flow to a certain place and must be treated to avoid harm to the environment or urban residents. It is therefore essential to ensure that waste is properly disposed of. In order to properly dispose of waste, it is absolutely necessary to first determine what kind of waste it is.

Disposal capacity management

Typical strategies for waste disposal include recycling, reducing volume or converting it into energy and dumping it into a landfill, or treating it through an incinerator, and then dumping non incinerable waste into a landfill.

Using our application and back-end ml analysis, incineration centers and landfills can manage capacity on an hourly basis, track each activity using dashboards, and map daily output and landfill usage. This allows them to plan accurately by using the combustion process wisely, the amount of flue gas used, and knowing the power required.

The number of collected by-products of flue gas, generated ash and unburned inorganic component solid blocks is also plotted. Appropriate actions can be taken in cooperation with other departments of the municipal government before the waste reaches the incineration center.

Pyrolysis is essentially the thermal decomposition of solid waste by applying heat without adding additional air or oxygen to produce by-products hydrogen, methane, carbon monoxide, tar and other inert materials. The weight of these by-products is also tracked to ensure that they do not pose a significant hazard to health and the environment. The by-products from the incineration center include low-grade concrete, which is then sold for bricks and other building and manufacturing blocks in accordance with the specifications established by the government authorities.

The capacity management and final treatment process can be drawn up very early, starting from the waste picture stage of the command center.

Add value to traditional process management

Historically, the term capacity management used to mean “managing various inventories in manufacturing plants” or “correctly determining the size of internal service delivery to meet current and future business objectives.” This is a kind of process management. In practical use, legacy systems combine external factors, such as product availability, market vitality, demand forecasting and internal resource allocation.

However, the ML waste management platform based on the Internet of things increases the scope of agile, real-time mapping and tracking through intelligent use technology, network, device or sensor management and machine automation. The integration of all these functions requires little manual intervention. Most activities are automated and monitored by an intelligent machine around the clock. The intelligent machine can analyze graphic data and perform some digital operations when needed.

Leave a Reply

Your email address will not be published. Required fields are marked *