- Demonstrator
AI Candy Machine
Automated sweet machine
The AI Candy Machine Demonstrator combines image processing and artificial intelligence. Originally developed as a clever sweet machine, the system recognises various objects using specially trained datasets.
The system is capable of both selecting individual sweets and recognising and sorting packets. It therefore also serves as a model for sustainable sorting and recycling processes. Development is taking place step by step: following the creation of our own datasets, the training of the recognition models and the development of initial prototypes, the system is being continuously expanded.
The aim is to turn a playful idea into a practical application that combines learning, automation and sustainability.
Highlights
Tools and technologies used
FAQ
What is the AI Candy Machine Demonstrator?
The AI Candy Machine Demonstrator is an interactive sweet machine that combines image processing and artificial intelligence. Visitors use a chatbot to select their preferred snack, which is then identified and classified by a trained image recognition AI before a robotic arm precisely picks it up and dispenses it.
How does object recognition work?
The system is based on specially trained datasets: once these datasets have been created and the recognition models have been trained, the AI can select individual sweets and recognise and sort packaging.
What does the demonstrator have to do with sustainability?
Although the app started out as a light-hearted game based on a sweet machine, the underlying sorting and recognition principle serves as a model for sustainable sorting and recycling processes – such as the automated separation of materials or packaging.
Who is developing the demonstrator?
The entire demonstrator is being developed and gradually expanded by an interdisciplinary team of students – from the initial data sets through model training to fully functional prototypes.