This project is the Stage 2 for my Smart Trashbin Project Series. Check out Smart Sort Kiosk for Stage 1.
Have you ever wonder which recycling bin should you use? With a paper coffee cup in your hand and standing in front of a huge recycling bin and wondering which compartment the cap goes and which hole should the cup go. Ever since the responsibility of recycling falling on consumers’ shoulders, it is never easy to do the job on a daily basis. There are too much trash to deal with and there’s no simple way to be environmentally friendly! You just not want to think every single time you want to get rid of anything. But you have to because you might not want to be the one to be blamed when you throw a recyclable can in the trash. Sadly, with all of these voices in your mind, you still do not know where to throw this paper cup away. When you laid your eyes on OmniRecycle, you know that your days of confusing recycling is over. This is going to relieve your mind from recycling forever.
The Background
The OmniRecycle project started when I was working on an AI project. When I finished my meal in a local sushi restaurant, I found that the employee of the restaurant mixed up the recyclables and the trash altogether. It just frustrated me because being an environmental advocate, I still did not want to give away to the convenience of using paper bowls and plastic lids. It was frustrating to see all my recycling effort went in vain when all the trash is mixed up. Thinking about the opportunities to improve this, I asked the question: can I make machines to do the recycling for me? That’s how I started my work on a smart trash bin. It turned out that with a simple image recognition model, the AI-equipped camera can be more reliable than humans to do trash classification.
Concept Description
OmniRecycle is designed to create a simple and pleasant experience of recycling.The flowchart shows the basic workflow of how trash get categorized.
The workflow of OmniRecycle When you put trash in the sorting platform, the sensor will detect that there is trash on the platform and the camera module will be evoked to take a photo of the trash on the platform. Then, the AI module will run the classification algorithm on the picture taken, and classify the type of trash. Based on the category of the waste, the motor control will turn the platform into the correct direction to drop the trash into the bin. The whole process is done automatically and painlessly. Cardboard Prototype Demo
A proof of concept demo for OmniRecycle
Product Launch Roadmap
To launch this product to the market, here is my plan for launching.
- Customer Discovery and Value proposition (week 1 — week 12) During this stage, I need to discover the customers and find the customer segments that have the needs for the technology. I need to form a few hypotheses of why they need the product and what form of product they want and test these hypotheses out.
- Product Design and Business Model (week 13 — week 24) Based on the tested hypothesis, I need to redesign the form and functionality of the product. Including advancing technology and improving the form of the product. In addition to the product design, I need to carry out a go-to-market strategy along with a revenue-generating business model. I need to carry out the
- Technology improvements (week 13–week 36) The technology improvements will be done in an iterative way along with the product design. Before the final design, I will fast prototype both the form and the software, then test them out in the real world scenario. Data collection and model training is a huge part of this process because the product’s performance largely relies on the accuracy of the model. User experience is also a huge part because we want the process to be pleasant and meaningful to the users. Also, we should start to develop our standardized testing procedure to ensure the quality of the product and the safety of the usage.
- Design for Manufacturing(week 24–week 36) By the end of week 36, I will have a final manufacturable design. The product should meet all the functions and design requirements and be ready for manufacture.
- Sale and Channel(Week 18 — week 24) Sale and Channel will begin at the 24 when we confirm all the customer needs and tested all the hypotheses. When we know what customer wants and what we can build, it is the time for us to sell it. Although the manufacturable design is not ready yet, we need to put our efforts into selling the product.
- QA and Delivery (Week 36–53) At the end of year 1, we should be able to have a minimum viable product that meets all the customer requirements and be ready to deliver to the customers. The product should pass all the stress testings that are defined during the product development process.