Cities Can Cut Plastic Waste Emissions by 96% With AI-Guided Strategy
Researchers have developed an AI framework that helps cities optimize plastic waste management by identifying which recycling methods deliver the biggest emissions cuts and economic returns. The study, published in Engineering, shows that cities following an optimal combination of interventions could reduce annual greenhouse gas emissions by 96.3% by 2060 compared to baseline practices.
The challenge is straightforward: cities lack reliable data on plastic flows, composition, and treatment outcomes. Without accurate information, managers cannot assess whether recycling, incineration, or landfilling makes sense for their specific situation. Machine learning fills these gaps by processing field measurements and cross-checking them against multiple data sources to catch errors and fill missing information.
Mechanical Recycling Wins in the Near Term
The framework evaluated different pathways for handling municipal plastic waste. Mechanical recycling - breaking down used plastics into raw materials - emerged as the most cost-effective option for the next 10-15 years. It produces emissions at roughly 108 kg CO2-equivalent per ton and generates economic returns around 613.9 Chinese yuan per ton.
Chemical recycling, which breaks plastics down to molecular components, shows promise but remains more expensive and less proven at scale. The researchers recommend treating it as a longer-term investment through demonstration projects rather than rushing it into widespread use.
Source Reduction Matters More Than High Recycling Rates
An unexpected finding: cities obsessing over recycling rates may miss bigger emissions wins. The optimal scenario combines source reduction (making less plastic) and bio-based substitutes with high-quality recycling. Over 30 years, this approach would prevent 22.22 million tons of CO2-equivalent emissions and generate approximately 197.7 billion yuan in economic benefits.
The researchers caution policymakers against treating recycling as a silver bullet. Pushing recycling rates higher without reducing plastic consumption can undermine long-term climate goals. Instead, they recommend treating source reduction and circular design as permanent policy constraints.
What This Means for City Planners
The framework provides actionable guidance for AI for Management professionals overseeing waste systems. It helps determine where to build recycling facilities, how to allocate budgets, and which technologies warrant investment. The AI approach works even when cities lack perfect data - a common situation in developing regions.
The methodology transfers across cities because it accounts for local conditions: waste composition, population density, economic activity, and existing infrastructure. A city can plug in its own measurements and get tailored recommendations rather than copying another city's strategy.
For operations teams, the framework clarifies the economics of different treatment pathways and quantifies trade-offs between emissions reduction and cost. This supports AI for Operations decisions about facility upgrades and process improvements.
The full study is available open access at https://doi.org/10.1016/j.eng.2026.03.009.
Your membership also unlocks:
