Although I criticize the AI bubble and the waste of energy by data centers on Earth, the technology itself is neither right nor wrong. Using it in the right place and applying AI to meaningful projects is valuable. When I see such products, especially hardware-related ones, I take note of them.Today, I want to share a Picky trash can, which is as follows:
This demo aims to analyze the type of trash being disposed of through visual analysis and provide different facial expressions to help improve people’s understanding of the confusion between recyclable and non-recyclable waste.
Specifically, take a look at this video:A sideways glance indicates incorrect disposal, while multiple nods indicate correct classification.
01
Functionality and Technical ImplementationThe functionality is straightforward, and the technical implementation principle is not complex, consisting of 4 servo motors and 3D printed structural components. Through a trained machine learning algorithm, it connects to the cloud to analyze and recognize the images of the trash, sending real-time signals to the Arduino based on the feedback, which controls the facial expressions of the doll.
02
Extended Value and SignificanceIn fact, this demo can enhance the functionality of the trash can’s electric door, first identifying the type of trash, then controlling the opening of the corresponding electric door, allowing only the correct disposal to occur. This not only judges but also achieves correct classification operations.On one hand, each time trash is disposed of, it serves as an educational moment for classification;on the other hand, it helps reduce the labor costs of manual sorting at community waste stations.I have written about this kind of solution before, especially during the phase three years ago when waste classification was widely implemented. We cannot ensure that everyone has a good awareness and knowledge of waste classification, but machines can. This is the value of robots; they should not just replace human actions but accomplish tasks that humans cannot do.AI accelerates this intelligent behavior.
03
Core Elements and Replicable ValueThis trash can project consists of two elements:AI through visual perception + combined with experience data from niche fields = a typical case of correct decision-making to solve pain points.Based on AI visual perception analysis, we can create countless valuable AI hardware simply by changing the experience data from a specific niche field.