The problem Smart Fridge solves

Fridge Talks addresses the significant global challenge of food waste, particularly within household refrigerators. According to the FAO, approximately one-third of all food produced for human consumption is lost or wasted globally each year, amounting to about 1.3 billion tons of food. In developed countries, 47% of this waste occurs at the consumer level, primarily within household refrigerators. This leads to substantial economic losses and environmental harm.

Fridge Talks is designed to mitigate this issue by using cutting-edge AI technology for smart food management and zero-waste living. The AI-Powered Fridge Manager offers the following benefits:

• Real-Time Detection: The system uses cameras to detect items being added or removed from the fridge in real-time. Advanced AI models analyze these images to recognize the category, quality, and quantity of each item, ensuring up-to-date inventory tracking.
• Interactive WhatsApp Bot: Seamless communication with the bot allows users to inquire about fridge contents, receive meal planning suggestions, and get alerts when food is nearing its expiration date.
• Food Spoilage Alerts: Notifications help ensure that users consume food before it spoils, reducing waste and saving money.
• Custom Meal Planning: Tailored meal plans help maintain a balanced diet while efficiently managing food inventory.
• Automatic Restocking Suggestions: Based on user preferences and consumption patterns, the system suggests items that need to be restocked to avoid running out of essential ingredients.

This solution not only reduces food waste but also enhances the overall food management experience, making existing tasks easier and safer. It helps households save money, improves health and nutrition by ensuring the utilization of available food resources, and contributes to environmental sustainability by conserving resources used in food production.

Challenges we ran into

While developing Fridge Talks, we encountered several challenges, particularly with the integration of various hardware and software components to achieve seamless functionality. One specific hurdle was ensuring accurate food item recognition and spoilage detection.

AI Model Integration: Initially, the AI models powered by TensorFlow and Mask R-CNN faced issues with accurately identifying and segmenting different food items in the refrigerator images due to variability in lighting conditions, item placements, and overlapping objects. To overcome this, we fine-tuned the ResNet model on a diverse dataset collected from Kaggle, which included images of common household food items labeled for supervised learning. Additionally, we implemented image preprocessing techniques using OpenCV to enhance the quality of the input images, improving the accuracy of item recognition.

Data Management and Communication: Managing the large volume of data generated by the AI analysis and ensuring efficient communication between the cloud server and the user interface was also a significant challenge. We used Django to develop a robust backend server that handles data storage and management. The SQLite database was used to store detailed information about the identified food items, user preferences, and system logs. Additionally, we created RESTful APIs to facilitate seamless data communication between the cloud and the user interface.

By addressing these challenges, we successfully developed a reliable and efficient AI-powered fridge management system that significantly reduces food waste and enhances the user experience.