Page 207 - AI for Good Innovate for Impact

AI for Good Innovate for Impact

schedules for those responsible for waste collection. This represents an innovative case that
demonstrates the potential of AI to realize resource circularity in public spaces, where it was
previously considered virtually impossible.

Once recyclables are separated from general waste, the next step is to sort them precisely Change 4.2-Climate
according to their intended use. For this, classification according to national or regional
recycling regulations is required. However, due to differences in regulations across countries,
localized adaptation of AI models is necessary. We developed the WEE (Waste Evaluation
Engine) algorithm package to enable rapid and cost-effective application of local recycling
standards. By identifying the needs of the Canadian container deposit refund market – which
had not been able to adopt AI due to complex regulations – we applied WEE and succeeded
in building a container-sorting AI system achieving over 98% accuracy within just three months.
(STEP II)

Lastly, beyond fine classification, we recognize the need for precision separation by material
composition. Currently, WEE is capable of most fine-grained classifications. However, we
are developing an advanced plastic composition classification system to overcome the final
challenge of identifying material compositions at a granular level, with the goal of launching
a commercial system within the next year.

WEE has been developed to flexibly adapt to regulations by country or intended use. Our
system is designed to quickly and accurately respond to any regulatory framework worldwide.
We believe that if our system is widely adopted globally, achieving a recycling rate of 60% in
the short term and over 80% in the long term is fully possible- along with improvements in
recycled material quality and sustainability.

This solution prevents municipal waste from entering the environment through efficient
treatment and recycling, thereby reducing water and soil pollution and protecting ecosystems,
improves urban waste management and air quality, reduce GHG emissions from transport
and treatment, and enhance infrastructure sustainability and eco-friendly urban development,
minimizes waste generation and promote reuse through well-separated recyclables. Optimize
plastic waste sorting to increase recycled plastic usage and support circular economy goals,
and reduces land-based marine pollution by using automated sorting systems to keep plastics
out of oceans, preserving marine ecosystems and enabling sustainable ocean use. Higher
recycling rates mean less landfill and incineration, reduced transport emissions, and more
circular resource flow. This benefits air, soil, and marine environments. The Busan Facilities
Corporation project proved that recycling is possible in previously unsorted areas, resulting
in higher recycling rates and better-quality secondary materials.

Use Case Status: Pilot

Partners:

Partner I: Busan Facilities Corporation [7]

Busan Facilities Corporation identifies and validates the value generated by our system. Installed
at Busan's busiest location, the organization worked with us to resolve real-world challenges
and stabilize operations. Even with many unfamiliar users, citizens became proficient in using
the system after about one year.

171