We were not able to create an actual smart recycling bin due to cost and resource constraints. Since we could not create an actual smart recycling bin, the performance of the model and solution has not been validated in a real-world setting.
- Streamlined sorting processes as workers no longer have to manually sort recyclables into Paper, Plastic, Glass, and Metal.
- Workers are only required to separate the items that cannot be recycled from the respective waste streams.
- Reduced contamination of recycled waste as waste is sorted into respective bins by smart recycling bin.
- Increased proportion of recyclable waste collected that is successfully recycled.
- Increased recycling rates.
A smart bin that utilizes deep learning models may require a large amount of power, which is not an aspect that was explored in this project. Ensuring the security and resilience of the deep learning models is crucial to maintaining the performance of the model. It could be susceptible to attacks by adversaries, such as malware attacks.
- Creating an actual physical prototype which includes integrating the model with the required hardware, sensors and testing in real-world conditions. Leveraging Raspberry Pis for container hosting also provides us with a distinct spatial efficiency advantage, due to the compact form factor of these units.
- Enhancing the functionalities of the bin by integrating sensors for detecting factors like waste level.
- Developing a mobile application to complement the smart recycling bin - waste management companies can receive alerts when the bin is full.
- Continuing to refine the deep learning model to enhance accuracy and speed.
In conclusion, our project embarked on a mission to revolutionize waste recycling and sorting processes by harnessing the capabilities of YOLOv8, a cutting-edge deep learning model. Although we faced constraints that prevented us from physically implementing the envisioned smart recycling bin, we successfully pivoted to develop a website that aims to simulate its functionality.
Our primary goal was to enhance the efficiency of recyclable waste sorting processes and increase recycling rates by addressing the pervasive issue of contamination in recycled waste. Through the development of our simulated system, we have taken significant strides toward achieving this goal.
By training our deep learning model to identify plastics, metals, paper, and glass, we demonstrated the potential for AI to play a pivotal role in the recycling industry. Our prototype serves as a testament to the innovative solutions that can drive environmental sustainability.
In a world where environmental conservation is paramount, our project represents a significant step forward in reshaping waste management practices. We are excited about the possibilities that lie ahead and remain dedicated to our mission of fostering responsible waste recycling and contributing to a greener, more sustainable planet. Together, we can continue to make strides toward a cleaner and greener future.