How does it work?
Introducing the AI Waste Sorting System, which predicts waste categories based on images and weight data. The front and top camera estimates the volume of waste, which helps us calculate the density of waste to complement Convolutional Neural Network image classification model. Powered by an edge device such as Raspberry Pi, fine tuning of the machine learning model is possible with cloud connectivity. There are a lot more cheap sensors that can be integrated to make more accurate predictions, and mass sorting the huge amount of data collected can be improved tremendously with deep learning.
How it would be produced?
As shown in the illustration, the Waste Sorting System is designed to be low cost, easy to assemble. The frame is made of PVC pipes and wiring of electrical components can be stored in the hollow part to avoid expose to weather or damage. The body is made of any lightweight sheet material (eg: acrylic). It has 2 detachable board platforms for easy maintenance and cleaning. There are no gears or bearings used. Servomotors are directly stuck into specially cut shaft end. The waste sorter is a light, modular design.
What makes it innovative?
The AI Waste Sorting System is innovative in the sense that helps companies, organization and government to understand consumption patterns within urban cities. Again, there are time series data that can be fed into the prediction model. For example, nearing lunchtime there may be more food waste generated at a local shop. With the integration of an NFC sensor, the public can choose to participate in recycling programs, exchanging their personal waste data for incentives streamed from targeted advertising or simply sales of anonymous/federated data. As 5G is coming into the picture soon, smart cities can benefit from the symbiosis of such an extensive network of physical data collected with production and sales sector.
Potential impact around the globe?
In a report by the World Bank, solid waste around the world is expected to increase by more than twice the human population growth rate by 2050. Poorly managed waste is one of the contributing factors of microplastic in oceans and our freshwater system. It is also notorious for causing pollution such as leachate and bad smell at landfills, bringing health and even life endangering problems to workers dealing with Municipal Source Waste. Since China banned imported waste since 2018, more countries in the Southeast Asia region started to restrict the sudden influx of waste.
Urban cities, especially those in developed nations, need to rethink their approach towards waste management and how to move towards a circular economy. What is the best way to counter these industrial speed production of complex waste? Could the barrier between waste and recycling simply be solved by effective sorting at source, before it avalanches to a seemingly unsolvable environmental problem? We believe the answer is YES! Only with proper sorting and building a reliable waste database can we tackle waste management realistically and formulate a system level strategy.
TOWARDS A ZERO WASTE FUTURE, WHERE THERE IS NO FINAL DESTINATION FOR PRODUCTS, BUT ONLY CHANGE OF MATERIAL FORM, JUST LIKE ENERGY.
1) REPLACING HUMANS IN SORTING WASTE MORE EFFICIENTLY.
2) SET UP A PROPER WASTE DATABASE TO ASSIST IN WASTE LAW RELATED ENFORCEMENT.