The RE-Team

Our project will tackle many issues, including pollution, human labor, and waste of energy. Our solution is that we will have two robots, one in the recycling factory, and the other will be deployed in public places as a form of a robotic recycling bin. The factory robot, built as an arm, will be classifying the garbage that will be put on the conveyor belt, whereas the public robotic bin will encourage people to throw the recyclable garbage in the right cabinets. The public bin robot will have three cabinets which are for collecting plastic, glass, and paper. Whenever a person throws garbage in the right cabinet, the robot will be thanking the person by basically saying "Thank You!" and conveying a happy animation. Whereas when a person throws the garbage in the wrong cabinet the robot will play a “Noo!” audio and illustrate a sad animation. The sanitation workers will have an update on the state of those robots to know if they’re full or not. Once they are full, they collect the cabinets and send them to the factory, where they will be put on the conveyor belt for the factory robot. We then set the robot in the factory to train on the new garbage data that has been collected by the public robotic bin. For example, we set the factory robot that now will be training only on plastic data, the next step would be to take the cabinet that held plastic material and put it on the conveyor belt. The factory robot will view the plastic from the top view and train itself to know that what is seen is plastic, so when it sees this material again it will detect it as plastic. The model will also be updated on the new different kinds of garbage, as time goes by different wrappers and garbage are being added to the bins, we highlight this as we are aware that garbage does not always look the same. Once the cabinets are emptied, they are then taken back to the public robotic bin for future use. From that repeating process, we can always assure an accurate model running on the robot that functions in the recycling factory. To summarize, our project will always be: 1. Improving the Azure Custom Vision model 2. Collecting free labeled data from the public through the robotic bins 3. Introducing the model to a new type of garbage The public robot will not always have the correct classified garbage as we are aware that not everyone will put their garbage in the right cabinet, but this error can be a good reference point to know how accurate our model is. Also, having a robot that automatically categorizes the garbage will keep the sanitation workers safe from being exposed to unhealthy chemicals while also having a more efficient method of recycling. While having a robot that encourages the public to throw their garbage will also assist in forming a healthier country. Our project saves energy by preventing the need of creating materials from scratch, but just simply recycling them. The main functionality of our project is to simply detect and categorize the garbage that will then be recycled and at the same time encourage the public to throw the garbage in the right dedicated cabinets. Our solution will not take over the jobs of the sanitation workers as they can still take the cabinets and empty them in the recycling company and later place back the cabinets easily on the robotic bin. Each cabinet will have a symbol indicating which cabinet stores which type of garbage.

We all received an email regarding the competition, thanks to Zaka, and none of us hesitated to join! We believe that it is a great opportunity for us to improve our skills in AI and ML and get the honor to work on a project for the Microsoft Imagine Cup. We were never in a Hackathon together, but interesting enough, we discovered that we all come from the same university, Qatar University. After the QIC Ideation day, we managed to form a group together and discovered that we all are interested in the Earth Field. With that interest, we were able to come up with an idea that we believe can make our planet healthier! Team Members: Azzam Alnatsheh, Qatar University, 2023, A computer engineering student who is passionate about Machine Learning and Artificial Intelligence. Abdelaziz Alswiti, Qatar University, 2024, is an electrical engineering student who is interested in developing electrical systems for various applications. Malak Sultan, Qatar University, 2025, An industrial and system engineering student who likes to gain new skills from different fields. Sara Abdel-Hadi, Qatar University, 2025, An electrical engineering student who is interested in solving issues by creating a smart systemic solution. Mentors: - Mr. Christophe Zoghbi - Mr. Christophe helped us with putting our idea together, as he helped us to get some inspiration and collect more information from different resources. Moreover, he always advises us to have a clear idea that relates to our reality - Mr. Joseph Assaker - Mr. Joseph Assaker aided us in improving our project by giving us a plan on how to proceed with the project in preparation for the round 2 submission date. In addition to that, he also proposed our ideas on how to create the necessary software that we need to present and submit.