Building a Real-Time AR Sorting System for ID Logistics

A sorting mistake in cross-border logistics can send a parcel to the wrong carrier or destination country, causing additional handling, unnecessary transport and delays. At ID Logistics’ distribution centre in Psary, operators had to match dozens of destination codes with physical pre-sorting fields – a process that depended heavily on experience and memory.

Nsflow, a NeuroSYS company, designed and developed Sorting AR: a dedicated system combining industrial smart glasses, computer vision, voice interaction, route optimisation and real-time integration with the customer’s Warehouse Management System. The pilot covers two sorting lines, 122 put-away fields and destinations across 23 countries.

Business challenge

The existing process identified which parcel was being handled, but it did not verify where the parcel was physically placed. An operator could scan the correct label and still select the wrong field, with the error becoming visible only later in the logistics chain.

The solution had to connect digital order data with a changing physical warehouse layout. It needed to identify parcels, determine their destinations, guide operators to the right location and confirm the actual put-away – without occupying their hands or slowing down the process.

The system also had to exchange data directly with eWMS, support several devices and sorting lines, and remain usable during temporary connectivity problems on the warehouse floor.

Because operators may distribute several parcels during one route, the application needed to optimise movement through the sorting area and update the sequence dynamically when another parcel was added.

1. Connect orders with physical locations

Current sorting orders are transferred directly from the customer’s eWMS. The platform maps parcels and destination codes to specific fields on the correct sorting line, while each device retrieves incremental updates and keeps the data required for continued operation during a network interruption.

2. Guide the operator and verify the action

After scanning a parcel, the glasses communicate its destination by voice and indicate the field on a map or directly in the camera view. Unique ArUco markers on hanging signs and floor labels allow the application to recognise the operator’s actual location and confirm whether the parcel has reached the correct field.

3. Optimise and protect the workflow

The application calculates an efficient route for several parcels and updates it as new items are added. It also records completed put-aways across device restarts, warns against repeated handling and detects parcels assigned to another line. Warehouse staff can reconfigure the zone in the web panel without changing the application.

Business impact

Sorting AR transfers knowledge about the sorting-zone layout from the operator’s memory into the system. Employees receive guidance and verification at the point where an incorrect decision could otherwise occur. This reduces the risk of put-away errors, supports faster onboarding and eliminates manual exchange of sorting data. The configurable architecture allows ID Logistics to adapt layouts and add further lines or locations without rebuilding the solution. Together with Inventory AR, the project also demonstrates how one platform and device environment can support multiple hands-free warehouse processes.

Client: ID Logistics – international contract logistics group (Psary distribution centre, Poland)

Industry: Logistics and warehouse operations

Scope of work: Process analysis, smart-glasses application, AR and computer vision, voice interaction, route optimisation, central management panel, offline synchronisation and eWMS integration

Technology: RealWear Arc 3, Unity, Android, Node.js, React, PostgreSQL and Kubernetes

Status: Operational pilot, 2026