Key results
The company
Bower
getbower.comMobile app for waste sorting and recycling rewards.
Result highlights
- 13M kg CO2 emissions saved
The story
A Swedish sustainability platform rewards users for recycling consumer packaging, aiming to apply the deposit-refund model to all waste types regardless of location.
Reliance on barcode scanning prevented global scale, as the system failed to recognize unfamiliar international packaging or items without identifiers. Furthermore, the manual effort required to map local waste infrastructure and drop-off points in each new country was prohibitive.
The company deployed a visual recognition system using Vertex AI to identify waste items from photos, extracting material data and brand information to calculate rewards. Gemini analyzes unstructured user feedback to identify valid recycling drop-off points, building a global knowledge base of local infrastructure. A continuous learning loop incorporates user validation of model predictions to refine accuracy over time.
Scope & timeline
- Expansion to 170+ countries in 6 months
Quotes
“Google Cloud simplifies the development and scaling of any kind of machine learning model. You don’t need ML expertise, you can just use its pre-built model library, Model Garden on Vertex AI, feed it data, and serve it at the click of a button.”