Sorting systems are judged by throughput and accuracy, but both are decided upstream of the sorter. The induction section, the scan tunnel and the merge are where parcels are either presented correctly or lost.
Key takeaways
- Design the induction for the parcel mix, not the average parcel.
- Control the gap between parcels before the scan tunnel.
- Provide a bypass or a recirculation for no-reads.
- Size the system for peak hour, not average day.
Parcel mix and induction
The parcel mix determines the conveyor width, the belt type and the induction method. A system designed for 300 mm cartons will struggle with 800 mm soft bags. Measure the actual mix: length, width, height, weight and surface friction.
Manual induction is flexible but slow and inconsistent. Automatic induction with a singulator is faster and more consistent, but it needs a parcel mix that fits the machine. For a mixed operation, a combination of manual induction for irregular parcels and automatic induction for regular cartons is common.
Gap control and scanning
The scan tunnel needs a gap between parcels so that each barcode is read once. A gap of 100–200 mm is typical, depending on the scanner and the belt speed. If the gap is too small, two parcels are read as one; too large, the throughput drops.
Use a pre-scan conveyor with a speed control and a photo-eye to create the gap. The gap should be created before the scan tunnel, not inside it. A parcel that is too close to the previous one should be held back and released with a gap.
| Section | Function | Key parameter |
|---|---|---|
| Induction | Singulate and orient | Parcel mix, rate |
| Pre-scan | Create gap | Gap 100–200 mm |
| Scan tunnel | Read barcode | Read rate >99.5% |
| Sorter | Divert to chute | Divert time, chute capacity |
| Chute | Accumulate and bag | Chute full sensor |
No-reads and recirculation
No-reads are inevitable. The system should route them to a recirculation loop or a manual encoding station, not to a random chute. A recirculation loop adds cost and space, but it protects the accuracy of the whole system.
Track the no-read rate by shift and by scanner. A sudden increase usually means a dirty scan window, a worn belt or a change in label quality.
Merge and diverts
A merge combines two or more infeed lines into one sorter. The merge must give priority to the main line and create a gap for the secondary line. A merge that simply pushes parcels together creates a jam at the sorter.
Diverts should be selected by parcel size and speed. A paddle divert is simple but can damage soft parcels; a pop-up wheel or a sliding shoe is gentler and faster. The chute should be sized for the parcel and should have a full sensor.
Peak capacity
E-commerce peaks are not average days. Design for the peak hour with a margin, and test the system at peak rate with representative parcels. A system that runs at 80% of rated speed in the test will not reach rated speed in production.
Provide buffer capacity before the sorter and after the chutes. A buffer before the sorter absorbs induction variation; a buffer after the chutes absorbs packing delays.
Frequently asked questions
What read rate should I expect?
A well-designed system with good labels should achieve 99.5% or better. Lower rates usually indicate gap control, label quality or scanner maintenance issues.
How much space does a recirculation loop need?
It depends on the sorter layout, but it is usually 10–20% of the sorter footprint. Plan for it at the design stage; retrofitting is difficult.
Can the system handle soft bags?
Yes, with a suitable belt, guide and divert. Soft bags are more likely to jam, so the conveyor width and the transfer gaps need to be designed for them.
Send your parcel mix and peak rate to our engineers through the contact page; we will help design the induction and conveyor sections.
