1. Home
  2. Case Studies
  3. Smart package sorting
Case study · eCommerce logistics

Smart package sorting: ML classification for an e-commerce logistics company

ScalaCode built a smart package sorting system for an e-commerce logistics company. A TensorFlow model classifies packages in real time from IoT-enabled conveyors and barcode scanners, and operators run the line from a Vue.js control interface.

IndustryeCommerce logistics
ProductAutomated package sorting system
Scala AIAITensorFlow package classification
HardwareIoT conveyor belts and barcode scanners
The client

An e-commerce logistics company

The client wanted faster package sorting with fewer manual errors. Packages had to be sorted automatically by destination, size and weight on its conveyor belts and scanners, on an architecture that could grow with daily volume.

The challenge

What made automated sorting hard

1

Real-time data volume

Scanners and sensors produce a constant stream of data that had to be processed in real time.

2

Many package types

The system had to be flexible enough to handle a wide range of package types.

3

Reliability at speed

Sorting runs at high speed, so the system had to stay reliable on a fast-moving line.

4

Manual handling errors

Manual sorting introduced errors the client wanted to remove.

The solution

Machine learning on the sorting line

A classification model, a high-speed backend and an operator console work together with the client's conveyors and scanners.

Sorting system settings for the ML model confidence threshold, database, notifications and security
  1. 01

    Package classification model

    A Python and TensorFlow model classifies each package in real time.

  2. 02

    Conveyor and scanner integration

    The system connects to IoT-enabled conveyor belts and barcode scanners.

  3. 03

    High-speed backend

    A Node.js backend processes the scan and sensor data, with PostgreSQL for storage.

  4. 04

    Operator control interface

    A Vue.js interface lets operators control the line and follow live operational data.

  5. 05

    Error and anomaly alerts

    The system detects errors and raises anomaly alerts for operators.

The results

Packages classified and sorted automatically

Packages are classified and sorted without manual handling, operators watch the line through live data visualisation, and anomaly alerts flag errors as they happen. The architecture is built to scale for large sorting operations.

Real-time

package classification

Automated

error and anomaly alerts

Tech stack

What Smart package sorting runs on

AI

Python, TensorFlow

IoT

IoT devices, conveyor belts, barcode scanners

Frontend

Vue.js

Backend and data

Node.js, PostgreSQL

Book a free strategy call

Plan a build like Smart package sorting for your business

  • 30 minutes with the ScalaCode team
  • What is realistic, what it costs and how long it takes
  • You decide afterwards whether to build with us

Book your strategy call

Takes 20 seconds