Page 211 - AI for Good Innovate for Impact

AI for Good Innovate for Impact

partner with capable local firms for such market entries and continue advancing our generative
AI and machine learning capabilities.

While AI is gaining traction in sectors like healthcare and automotive, its transformative potential
in recycling remains underutilized. Canada, for instance, faced sorting and storage crises due Change 4.2-Climate
to labour shortages—issues our AI helped resolve. The lack of AI application in recycling stems
from a lack of domain-specific knowledge. We offer both technical expertise and field-tested
know-how and are ready to collaborate with any partner globally. Beyond technological
innovation, finding capable partners is another important next step. By working with more
organizations that need us, we are one step closer to full circularity.

3 Use Case Requirements

REQ-01: It is required to have datasets being designed flexibly for reuse across different
contexts. The dataset from this solution is large-scale and structured for multiple uses as
recycling containers are globally similar in material—glass, plastic, paper, metal—but vary in
shape, size, and purpose.

REQ-02: It is required to have effective integration and fast processing as classification is
complex and requires multiple AI algorithms. The contamination detection must include
removable, washable, or non-recyclable pollutants while sorting may differ by country (e.g.,
Canada separates aluminum vs. bi-metal cans).

REQ-03: It is required to have a hardware system including recognition (for accurate detection),
transport (to move and sort), and management (to leverage data and generate services).

REQ-04: It is recommended to use WIMPLE-generated data as they can reduce costs by
optimizing bin replacement timing, route planning for collection vehicles, and recycling
logistics. Real-time carbon reduction metrics also support ESG reporting.

175