The setting
The Ministry of Hajj and Umrah's Hajj Without Luggage programme for 2026 moved pilgrim baggage from hotels, camps and city collection points through airport processing to hotel delivery, without the pilgrim carrying a bag. HalaBag's system ran the orchestration: approximately 1.2 million bags, seven airports live at once, multiple airlines, handlers and logistics providers, and a programme authority watching the numbers.
No ordinary season produces that concentration. Here is what it taught us about running baggage across many stations at once.
1. Allocation must be a rule, not a person
At peak, thousands of cases an hour needed a provider. Any process that depended on a coordinator choosing a courier would have collapsed by the second day. Allocation ran on rules: which providers cover the lane, what capacity they have declared, what their measured on-time rate is, and what the case costs on each. Operations staff handled exceptions and overrides, each with a logged reason. That is the only way allocation scales, and it is also the only way it becomes auditable.
2. Custody is confirmed by the receiver, or it is not custody
Every hand-off, from collection point to handler, handler to aircraft, aircraft to arrival handler, handler to provider, provider to hotel, was confirmed by the receiving party with time, location and evidence. Where a confirmation was missing, the case was flagged rather than assumed complete. Liability for a bag sits with whoever had custody; the chain has to be complete for that to mean anything.
3. Stations are configuration
Seven airports meant seven sets of cut-offs, handlers, providers and local rules. None of them was a project. Each station was configured on the platform: products, cut-offs, documents, steps, providers. That is the difference between a platform and a bespoke build, and it is why the second airline on a platform costs a fraction of the first.
4. Exceptions are the job
With allocation and custody systematised, the operations team's work became the exception queue: cases approaching their SLA, hand-offs not confirmed, addresses that did not resolve, providers not accepting. A queue that surfaces those before the deadline is the entire operational value of the layer. A dashboard that shows them afterwards is a report.
5. Repeatability is the proof, not the volume
The question every enterprise buyer asks about a seasonal operation is what happens the rest of the year. The answer is the same platform running the same providers on a national carrier's mishandled-baggage recovery every day, and a standard implementation model: tenant, stations, source-system connection, providers, go-live checks. The volume made the headline. The repeatability is what a buyer should look at.
What we would tell an airline planning multi-station rollout
- Start with one station and one product, with the source system connected.
- Integrate providers once, on one case model, so they are available at every station.
- Treat the exception queue as the operations console.
- Measure providers from custody events, not from their reports.
- Add stations by configuration and prove it by adding one without a developer.
Figures from HalaBag operations data for Hajj 2026; programme details from the Ministry of Hajj and Umrah and Saudia's published announcements.