Shipsy’s hub operations automation turns a manual sortation hub — the kind that relies on experienced supervisors, chalked zones on the floor, and daily firefighting — into a system of directed scans, predictive wave planning, and auto-generated bag manifests. Enterprises deploying the full stack typically see 2–3x sortation throughput and a substantial drop in misrouted parcels, without adding staff.

Why we built this

Hubs are the choke point. A national parcel network’s cost-per-shipment is set more by what happens between 10pm and 4am at three or four sortation hubs than by anything drivers do during the day. And hubs are chronically under-instrumented: a supervisor manages 50 staff against a chalked floor plan, decisions are shouted, and misroutes are caught at the next hub (or at the customer’s door).

Enterprises had either over-invested in fixed-sortation hardware that couldn’t flex, or under-invested in tooling and lived with 2–5% misroute rates. We built hub ops automation as a flexible software layer — the same system runs on a simple manual hub with tablets and a sub-million-shipment network as on a conveyorized hub with 2M+ shipments/day.

How it works

Four mechanisms compose the hub stack:

Mechanism 1 — Inbound wave planning. Pre-arrival, the system ingests the vehicle manifests of all inbound trips, groups shipments by destination, by service level, and by line haul trip they need to exit on, and generates a wave plan for the hub — which door each vehicle parks at, which zone it unloads into, which sort chute its shipments route to, and which outbound vehicle they load onto. Wave plans are published 30–60 minutes before vehicle arrival so staff is staged.

Mechanism 2 — Scan-directed sortation. Every shipment is scanned at induction. The system returns a sort decision — chute, zone, outbound trip, or hold — based on the live wave plan and dynamic capacity. Sort decisions update continuously: if a zone fills up, subsequent scans re-route to an overflow zone. Misroutes surface instantly — a scan that would cross-contaminate an outbound trip is rejected at the scanner with a clear audio/visual alert.

Mechanism 3 — Bag, LPN, and manifest generation. Shipments destined for the same outbound trip are consolidated into bags (for CEP/parcel) or LPN-labeled pallets (for B2B and freight). The system generates the bag/LPN identifier, prints the label, and attaches a digital manifest — a structured list of all shipments in the bag, their destination, weight, volume, and COD if any. At the outbound vehicle, a single bag scan manifests dozens or hundreds of shipments into the trip.

Mechanism 4 — Hub control tower. A live dashboard shows hub state at a glance: vehicles parked vs expected, inbound backlog by zone, sort chute utilization, outbound readiness by trip, staff deployment vs wave plan. Supervisors get exception alerts — a zone approaching capacity, a chute not moving, a trip with missing shipments — before the miss happens.

All of this runs on Shipsy’s middle-mile product, and integrates with the TMS for outbound trip planning and with the driver app for final-leg handoff.

Here’s the flow at a glance:

flowchart LR A[Inbound vehicle] --> B[Induction scan] B --> C[Sort decision] C --> D[Bag by destination] D --> E[Print LPN + manifest] E --> F[Outbound vehicle load] F --> G{Trip complete?} G -->|Yes| H[Release for linehaul] G -->|No| F

Early results

Enterprises deploying hub ops automation typically report, within 90 days:

A leading Western European parcel operator with 50%+ national market share uses this stack across its hub network to maintain 8M+ shipment throughput with tight cut-off windows.

What’s next

Three upgrades: conveyor-integrated sortation (direct control of automated diverts where conveyor infrastructure exists), predictive staffing (the wave plan proposes staff allocations per zone based on forecast volume), and cross-hub visibility — a network-level control tower that treats all hubs as one system, surfacing structural bottlenecks that a single-hub view can’t see.