Automated anomaly detection increases transparency in waste certification
The customer
The challenge
Cleanhub issues waste certificates to companies for the financing of waste management projects. To ensure trust and actual environmental success, it is necessary to accurately record data such as the weight of waste bags, image documentation, tracking of collection trucks and identification of households.
Detecting fraudulent activity manually is inefficient and difficult, which is why an automated anomaly detection system was needed to ensure data transparency and reliability.
Solution
We carried out a comprehensive data-driven project in several agile sprints. In the first phase, validity checks, data pre-processing and an exploratory data analysis (EDA) were carried out to gain initial insights.
In the following sprints, data augmentation, feature engineering and the development of anomaly detection models for registration and waste data were implemented. In parallel, we advised Cleanhub on optimizing their data collection methods and provided detailed recommendations for future development.
Results & effects
Cleanhub has significantly improved its ability to detect anomalies and suspicious activity. The increased data quality and automated detection now provide accurate, actionable insights and strengthen the credibility and reliability of their certification processes.
In addition, our strategic recommendations support Cleanhub in continuously developing its approach and effectively intensifying the fight against greenwashing.
Overview
- Waste management & certification
- Offers companies waste certificates to offset waste
- Supports local waste disposal companies in Southeast Asia
- Ensures transparency and compliance to prevent fraud and greenwashing
- Exploratory data analysis (EDA)
- Data validation and pre-processing
- Modeling for anomaly detection
- Data augmentation
- Strategic advice on data collection
.png)
.png)
