Eco-Impacto

A platform that turns the IoT infrastructure of smart bins (RFID readers and weight sensors) into citizen engagement and auditable ESG metrics for the council. Without installing any new hardware.

Context

The smart bins are already installed and already measuring, but they are used only for technical control: fill level, incidents and routes. The system is invisible to the citizen, who experiences recycling as an obligation rather than as participation. Meanwhile, the share of contaminants in the organic fraction degrades compost quality and forces additional processing.

Solution

The same telemetry, three views: a gamified citizen app with measured impact (CO₂ avoided, compost that can be generated, neighbourhood percentile), a municipal dashboard with purity, fraud and routes, and a technical layer that exposes the real IoT flow (payloads, latencies, anti-fraud mechanisms acting on the data). Every deposit is physically validated by sensor, so the recognition cannot be faked.

The AI inside

What runs on its own.

1. 01

   Anomaly detection

   Flags anomalous weight patterns and outlier frequencies against each user’s and bin’s history, with the evidence of the event that triggered the alert.

2. 02

   Route optimisation

   Reorders collection by real fill level instead of a fixed calendar: fewer kilometres per route and fewer empty pickups.

3. 03

   Translation into impact

   Turns kilos of organic waste into CO₂ avoided and compost that can be generated, with the same figure for the citizen and for the council’s ESG report.

Screenshots

• 
• 
• 

Indicators

• 3 Views on the same data: citizen, municipality and technical layer
• 0 New hardware: it runs on the existing IoT infrastructure
• 90 days Proposed pilot in a neighbourhood with bins already installed

What it solves

• <3% Openings with no deposit · pilot target against a 8–12% baseline
• <12% Contaminants in the organic fraction · target against the current 18–22%
• −26% Collections per bin per month · pilot target, from 8.2 to 6.1 with routes by real fill level

Process

How we solved it.

1. 01

   Diagnosis

   The operator’s two problems: waste purity and citizen adoption, on infrastructure that is already installed

2. 02

   Data model

   One sensor event as the single source: it feeds gamification, operations and reporting with no parallel recalculation

3. 03

   Prototype

   Three navigable views with coherent data: citizen, municipality and technical layer

4. 04

   Pilot

   Pilot neighbourhood, 90 days, hypotheses measured against the baseline before scaling

Same infrastructure. More layers of value.