<aside> 📽️
</aside>
<aside> ℹ️ Put the names of all team members under your video.
Zazi Moloi - https://www.linkedin.com/in/zazi-m-8a223367/?lipi=urn%3Ali%3Apage%3Ad_flagship3_feed%3BmSE0NGaSTYGVu%2FGDaw8auQ%3D%3D
Zongze Li - www.linkedin.com/in/zongze-li-625978301 Abhishek Niaranjan - https://www.linkedin.com/in/ahniranjan/
</aside>
[<150 words]
<aside> ℹ️ In Paraisópolis, a dense informal settlement in São Paulo, residents lack safe, nearby and reliable ways to dispose of household waste and recyclables. Even when waste is collected in the wider city, weak last-mile access and irregular service in high-density areas lead to overflowing points, contamination of recyclables, illegal dumping and burning. This creates public-health risks (smoke, pests, blocked drains), environmental harm, and lost recycling value. The problem also directly impacts informal waste pickers (catadores) and cooperatives, whose livelihoods depend on recoverable materials but are undermined by mixed/contaminated waste and unsafe collection conditions. Local businesses (markets, restaurants and shops) contribute significant waste volumes and are affected by hygiene and compliance issues when collection is unreliable.
<aside> ℹ️
The Paraisópolis Clean Loop tackles last-mile waste failure through five integrated components: simple three-stream source separation, secure micro-collection hubs placed within a 2–3 minute walk, GPS-verified pickup routes, formal catador integration, and commercial waste contracts.
The system is technically feasible — it requires no expensive sensors, relying instead on adaptive routing and low-cost monitoring. It is financially viable — commercial contracts stabilise revenue while recovering value from recyclables currently lost to landfill. It is socially acceptable — minimal burden on households, visible cleanliness improvements, and formalised, protected roles for waste pickers rather than displacing them. It is environmentally sustainable — diverting organics (~45% of waste) and dry recyclables directly reduces dumping, burning, and landfill pressure.
By addressing infrastructure gaps, behavioural barriers, and livelihood protection simultaneously, the Clean Loop delivers measurable impact for residents, catadores, and the municipality alike.
</aside>
[<150 words]
<aside> ℹ️
Several assumptions underpin this proposal. We assume residents will adopt three-stream sorting with sufficient consistency — participation rates in informal settlements are uncertain and context-dependent. Hub placement assumes community acceptance and safe siting, which requires local negotiation not yet completed. Catador cooperation assumes willingness to formalise, which may face resistance if trust is low or income fears arise. Commercial waste volumes and contract revenues are projected, not confirmed. Contamination rates and diversion potential are estimated from national data, not Paraisópolis-specific measurement. To address these uncertainties, Phase 1 is deliberately small-scale — 1–2 micro-zones, 3–5 hubs — functioning as a structured learning pilot. Weekly improvement loops, participation surveys, contamination checks, and GPS route audits generate real local evidence quickly. Expansion only proceeds zone-by-zone as performance data validates assumptions, ensuring the system adapts to ground truth rather than projections.
</aside>
<aside> ℹ️
Paraisópolis is São Paulo's second-largest favela, housing approximately 100,000 residents across a densely built, hilly terrain. Standard municipal collection vehicles cannot reliably access many internal streets, creating persistent service gaps. Where collection fails, residents resort to burning or dumping — generating smoke, pest proliferation, and blocked drainage that amplifies flood risk during São Paulo's heavy rain seasons. This is not unique negligence; it reflects a structural mismatch between formal waste infrastructure designed for grid-plan cities and the organic, high-density layout of informal settlements.
Brazil's national picture reinforces why this matters at scale. In 2022, ~81.8 million tonnes of MSW were generated nationally, with ~93% collected — yet ~39% of collected waste still went to dumps rather than sanitary landfills, and recycling recovery reached only ~1.47% (Municipal solid waste treatment in Brazil: A comprehensive review, n.d.). The bottleneck is not awareness or political will alone — it is last-mile execution and sorting infrastructure.
A critical insight shaping this proposal is that low participation in waste sorting is often misattributed to resident indifference. Evidence suggests otherwise: when infrastructure is inconvenient, unclear, or unreliable, rational actors disengage (Li et al., 2024). Contamination rises when sorting categories are confusing or hubs overflow. Catadores lose income when contamination devalues loads. Municipalities lose revenue when recyclables are landfilled — an estimated US$2.5 billion per year nationally (Recyclable waste in Brazilian municipalities…, 2023).
This creates a reinforcing failure loop:
Poor infrastructure → low participation → high contamination → low recycling value → reduced investment → poorer infrastructure
The Clean Loop is designed to interrupt this cycle at multiple points simultaneously.
| Barrier | Component | Evidence Base |
|---|---|---|
| Sorting confusion → contamination | Simple 3-stream system | Li et al. (2024) |
| Distance/inconvenience → non-use | Micro-hubs within 2–3 min walk | Recyclable waste… (2023) |
| Unreliable pickup → overflow/dumping | GPS-verified fixed-calendar routes | Spinelli et al. (2025) |
| Catador exclusion → system fragility | Formal integration with PPE + stable roles | Li et al. (2024) |
| Revenue instability → service gaps | Commercial waste contracts | Vehicle routing… (2022) |
The three-stream approach — wet/organic, dry recyclables, reject — is deliberately simple. Evidence from behavioural research in waste contexts shows that complexity is a primary driver of sorting errors and participation fatigue (Li et al., 2024). Plastics sub-sorting is deferred to Phase 3 when habits are established.
One underappreciated failure mode in urban waste systems is route unreliability. Vehicles fill before completing routes; hills or narrow lanes cause delays; missed streets accumulate. Spinelli et al. (2025) demonstrate that static routing underperforms in variable real-world conditions, and that rolling-horizon adaptive replanning significantly improves service reliability. The Clean Loop incorporates this by building a weekly replanning loop from GPS data — missed hubs trigger route adjustment the following week, not the following quarter.
Additionally, separating collection frequency by stream (organics collected more frequently due to health risk; dry recyclables 2–3 times per week) reflects operational best practice from hybrid vehicle routing research, which shows that matching vehicle capacity and route design to waste generation patterns reduces both cost and overflow incidents (Plug-in hybrid electric refuse vehicle routing…, 2022).
Brazil's informal waste sector is vast — hundreds of thousands of catadores contribute significantly to whatever recycling recovery Brazil currently achieves. Yet poorly designed "circular economy" transitions can inadvertently threaten these livelihoods when new systems bypass or replace informal workers rather than incorporating them (Li et al., 2024).
The Clean Loop treats catadores not as a legacy problem to be solved, but as a core operational asset:
This also strengthens community trust, which is essential for sustaining household participation over time.
The phased commercial onboarding strategy is central to long-term financial sustainability. Restaurants, hotels, markets, and small industries generate predictable, high-volume organic and packaging waste — and many face regulatory pressure to manage it responsibly. Separate commercial contracts provide:
This mirrors standard practice in municipal waste economics, where commercial accounts are essential to system viability (Plug-in hybrid electric refuse vehicle routing…, 2022).
The three-phase structure is not merely cautious — it is epistemically necessary. Assumptions about participation rates, contamination levels, catador uptake, and commercial volumes must be tested before scaling. Phase 1 (0–8 weeks, 1–2 micro-zones) generates the local evidence base. Phase 2 expands only where performance data supports it. Phase 3 adds organic processing infrastructure once diversion volumes justify the investment.
This avoids the common failure mode where well-designed pilots collapse at scale because early assumptions were never validated — a pattern consistent with Brazil's history of policy reform without effective local execution (Recyclable waste in Brazilian municipalities…, 2023).
By Month 12, a successful Clean Loop pilot would demonstrate:
These outcomes would constitute proof-of-concept sufficient to justify municipal partnership and zone-by-zone expansion across Paraisópolis — and a replicable model for other informal settlements across São Paulo and Brazil.
</aside>
<aside> ℹ️
Li, B. et al. (2024) 'Competing narratives inhibit a circular economy for bio-based plastic packaging: Insights from a social innovation lab study in Brazil, Canada, Poland and UK', Business Strategy and the Environment.
Municipal solid waste treatment in Brazil: A comprehensive review (n.d.)
Plug-in hybrid electric refuse vehicle routing problem for waste collection (2022) Transportation Research Part E: Logistics and Transportation Review.
Recyclable waste in Brazilian municipalities: A spatial-temporal analysis before and after the national policy on solid waste (2023) Journal of Cleaner Production.
Spinelli, A., Maggioni, F., Ramos, T.R.P., Barbosa-Póvoa, A.P. and Vigo, D. (2025) 'A rolling horizon heuristic approach for a multi-stage stochastic waste collection problem', European Journal of Operational Research.
</aside>
<aside> ℹ️ You are welcome to put additional information in an appendix!
We recommend making any appendix a subpage that is located here. You can reference that page at appropriate locations above using the “@” and the name of the page; for example: Example Appendix Page references the page located just below here.
</aside>