Client: NDA

Advanced AI System for Seamless Bottle Recycling

Case study details

A pioneering UK start-up, embarked on a bold mission to revolutionize the bottle recycling process. Recognizing the longstanding challenge of execution within the recycling industry, particularly in terms of user-friendliness, company sought to innovate with a solution that would streamline and enhance the user experience. Their vision? To introduce stand-alone recycling machines operational 24/7, designed to pick up bottles ready for recycling.

Challenge

Advanced AI Integration: Automated Recycling Machines with Material Recognition and Anti-Fraud Measures

The machines, equipped by default with computer vision sensors, were enhanced with an additional validation layer capable of identifying the material composition of bottles, assessing their size and condition, and—based on that information—rewarding users, referred to as depositors, with credits.

The main challenge involved building and then integrating an accurate and performative AI model that could recognize the objects that came into the machines, incorporate the anti-fraud systems and make it all work locally on the internal software infrastructure consisting of Raspberry PI.

Our team expert opinion

While the initial concept appeared straightforward to implement, the software posed significant constraints. We faced the challenge of constructing a model capable of operating within a very limited time and within relatively modest infrastructure. Unlike cloud-based solutions boasting substantial computing power, our hardware, based on Raspberry Pi, presented inherent limitations. Thus, our strategy necessitated seeking out compact AI models instead of relying on state-of-the-art computer vision solutions.

Michał Pocztowski Senior Data Scientist at Addepto

Approach

Multi-Stage Classification: Enhancing Recycling Accuracy and Preventing Fraudulent Attempts

The process had to be divided into several stages, with the first being a two-step classification. The sensors within the cameras, observing the items deposited by users, had to distinguish between glass, plastic, and cans. Additionally, in the case of plastic and cans, they had to assess their degree of damage.

This determination ultimately dictated the number of points awarded to the user. The system also had to recognize fraudulent attempts, meaning it had to differentiate organic waste and identify reverse movement of objects on the conveyor belt in case a user attempted to recycle the same item multiple times.

For the classification, we employed a pre-trained MobileNet model, and for assessing object sizes based on semantic segmentation, we utilized DeepLab version 3. Both models were fine-tuned with our data to tailor them to our specific use case. To detect reverse movement of objects on the conveyor belt, we use “motion vectors” provided by the Raspberry PI camera.

– Michał Pocztowski, Senior Data Scientist at Addepto.

Outcome

Advanced AI Systems for Seamless Bottle Recycling

The company’s stand-alone recycling machines enhance user experience by providing a convenient and accessible way to recycle bottles round-the-clock. The integration of advanced computer vision sensors ensures efficient material identification, allowing accurate sorting and processing of bottles.

Moreover, anti-fraud systems effectively differentiate organic waste and detect reverse movement of objects on the conveyor belt, promoting fair recycling practices.

Before

After

About Addepto

Addepto is a leading AI consulting company, recognized by Forbes and Deloitte, for delivering cutting-edge AI and Data-driven solutions. We specialize in accelerating process automation and optimization within global enterprises using modern technologies.

Here you can learn more about the technologies used in this project:

About us