blog banner picture

case study

Z-Enviro Industries | Case Study

Published On

2 minutes to read

Z-Enviro Industries Pvt. Ltd is one of the largest electronics recyclers in Hyderabad, with a recycling capacity of 10,000 tons per year. It was founded in 1999 as Famous Trading Company and mainly dealt with electronic scrap. The company also imported used computers from China, Hong Kong, Malaysia, and other countries.

Currently, Z Enviro, as a firm, is working at the pan-India level and provides complete end-life management of electronic equipment, providing a one-stop, cost-effective solution.

Opportunity:

The company issued invoices to LG, Samsung, etc. in order to process e-waste for its clients. The invoice must be accompanied by images demonstrating how they recycled e-waste from vehicles. Invoices were being rejected due to the submission of duplicate images in the past. It would be advantageous to avoid this so that invoices are cleared without delays and preserve its reputation with the client.

Approach:

We studied the problem in order to provide an optimized solution to Z Enviro. After the image is clicked through the app, the machine learning algorithm will compare the image with past 30-day images in the database for any duplication.

The image would be rejected if it was determined to be a duplicate. This happens at the granularity of each appliance in the Image.

Measurement Metric:

A 95% accuracy rate was achieved when matching images. The algorithm identifies not just at the image level, it identifies at the item level. Therefore, if a TV is present in both images but all other appliances in that image are different, the algorithm will also flag that image.

Results:

Fraud submissions have significantly reduced with the application, and invoices have been cleared without hiccups. Most importantly, Z-Enviro is well-positioned to get clients.

Related Blogs

Bringing together experience & innovation.