an aerial view of an airport
心を動かす会話

Transforming Stand Allocation with a Predictive Digital Twin

A platform that anticipates stand conflicts before they happen, and automatically optimizes the airport's full assignment plan, moving from static planning to predictive operations.
会議のスケジューリング
課題
Every assignment shapes the rest of the operation

Assigning a stand to each flight looks like an isolated decision, but it conditions everything that follows. A stand occupied longer than planned can block the pushback of the neighboring flight, leave an aircraft waiting, and trigger delays that ripple through the entire day.

187
Published stand positions
1
Full taxiway network modeled
47%
On-time departures, pre-optimization
100%
Planning still static and reactive
two airpot controllers working on a digital twin application in a computer
Static Planning
Stand plans are fixed in advance and don’t adjust when the real-time operation diverges from plan.
Domino Effect
A conflict at a single stand propagates to neighboring flights and contaminates the rest of the day.
Late Detection
Conflicts are discovered once they’re already happening, not before, leaving little room to react.
当社の取り組み
Three pillars of intelligent stand allocation

This isn’t about replacing the planner: it’s about giving them foresight, an automatic solution, and a reason to trust every change.

Predict — Don't react
Airport Digital Twin

We replay a full operating day on the airport’s real geometry; its stands, its taxiway network, its flights — and in seconds the system scans the entire day, flagging where each conflict appears and which flights it affects.

Optimize — Not just detect
Automatic Reassignment Under Real Constraints

The optimizer takes the operation’s real-world constraints, aircraft size, shared boarding bridges, common pavement and automatically generates a new stand assignment that eliminates the detected conflicts.

Explain and adapt — Not impose
Traceability and Minimal Re-optimization

Every change can be queried for its rationale. And when the real operation diverges from the plan — a flight is delayed, a stand goes out of service, the system re-optimizes, changing only what’s necessary: flights already boarding are left untouched.

アーキテクチャ
An optimizer that understands the real operation

A modular architecture that simulates the full operating day, detects conflicts, and continuously re-optimizes stand assignments around real-world constraints.

Predictive Digital Twin Approach
Simulates · Detects · Optimizes · Adapts
Predictive Digital Twin Approach
Simulates · Detects · Optimizes · Adapts
  • Simulation engine (digital twin)

    Models the airport’s exact geometry: 187 published stands and their taxiway network, replaying a full day in seconds.

    Simulation engine (digital twin)
  • Conflict detector

    Automatically identifies where and when each stand conflict occurs, and which flights are affected.

    Conflict detector
  • Reassignment optimizer

    Generates a new stand plan that respects real operational constraints (aircraft size, shared bridges, common pavement).

    Reassignment optimizer
  • Explainability engine

    Surfaces the rationale behind every proposed change.

    Explainability engine
  • Incremental re-optimizer

    Recalculates the plan around disruptions (delays, out-of-service stands), touching only what’s strictly necessary and preserving operations already underway.

    Incremental re-optimizer
Live Scenario
One day at Guarulhos: optimizing the plan from 47% to 100% On-Time Performance

1. Detection. The digital twin replays the full operating day and flags 89 stand assignment conflicts.
2. Optimization. The optimizer generates a new stand assignment, honoring real aircraft and boarding-bridge constraints.
3. Result. All 89 conflicts are eliminated, idle time is cut by nearly half, and the disrupted operation is redistributed across just over 50 flights, creating an optimized plan with 100% on-time performance.
4. Explainability. Every change is logged and available for review.
5. Live Disruption. A flight is delayed; the system re-optimizes only the affected plan, without touching flights already boarding.

Play the video below

Predictive optimization in action

-15
再生アイコン
+15
Detect conflicts

⚠️ See where conflicts happen — before they impact the operation.

Optimize the plan

🔧 Generate a better stand assignment around real operational constraints.

See the impact

✅ 89 conflicts eliminated, downtime nearly halved, 47% → 100% on-time departures.

Adapt to disruption

🤖 Change only what’s necessary as operations evolve.

00:00
/
00:00
ミュートアイコン
再生アイコン
「閉じる」アイコン
ドラッグ