top of page

Gate Assignment Tool

*Confidential Project – Summary Only

+

Overview

GateAI is a gate assignment optimization platform co-developed by Optym and Southwest Airlines. It automates and assists in real-time gate allocation by combining AI-driven recommendations with human-in-the-loop controls.

​

The goal: reduce planner workload, prevent gate conflicts, and improve throughput at some of the busiest airports in the world.

My Role & Collaboration

I was the UI/UX Designer on the team, responsible for end-to-end design:

  • Research (interviews, process mapping, comparative analysis).

  • Information architecture and workflows.

  • Wireframes, interactive prototypes, and micro-interactions.

  • Usability testing and iteration.

​

I collaborated closely with:

  • Gate planners & controllers (end users).

  • Product leads (business priorities).

  • Engineers & algorithm scientists (translating constraints into usable, explainable UI).

Research

Methods:

  • Stakeholder Interviews → gate planners, NOC controllers, GroundOps.

  • Process Mapping & Observations → documented workflows for normal vs. disruption-heavy operations.

  • Comparative Analysis → benchmarked existing ops tools and control center dashboards.

  • Design Sprints → facilitated workshops around conflict resolution, prediction visibility, and override mechanisms.

​

Findings:

  • Planners don’t trust black-box automation, AI rationale must be visible.

  • Stress peaks during irregular ops (delays, diversions, congestion).

  • Users wanted color-coded alerts, “what-if” tools, and historical usage patterns.

  • Override flexibility with minimal clicks was critical.

Synthesis & Insightslaboration

We synthesized research into planner workflows and needs:

 

Personas:

  • Station-level Gate Planner → focuses on immediate conflicts, crew/passenger connections.

  • NOC Supervisor → monitors the network view, balances system-wide priorities.

​

Workflow Pain Points:

  • Fragmented data sources → slowed decision-making.

  • Reactive firefighting instead of predictive planning.

  • Lack of collaborative visibility → redundant phone calls.

Ideation & Design Process

We moved from problem framing → idea generation → testing:

  • Early Sketches → paper flows of Gantt charts, conflict resolution, and AI explanation widgets.

  • Wireframes → timeline-based scheduling, gate assignment drag-and-drop, alert panels.

  • Interactive Prototypes → tested with planners to simulate both calm and disruption-heavy scenarios.

Prototyping & Testing

  • Tested interactive prototypes with Southwest planners.

  • Iterated on color coding for alerts until critical vs. low-priority conflicts were instantly distinguishable.

  • Refined sandbox mode after users asked for a safe testing space without impacting live ops.

  • Added explanation tooltips after feedback that “we need to know why the AI chose this gate.”

Final Solution

Core Design Features:

  • Gate Plan Visualization: Dual Gantt + Grid views; aircraft “pucks” with real-time progress bars and hover details.

  • Conflict Resolution: Drag-and-drop reassignment, side-by-side comparison of old vs. new assignments, batch conflict resolution.

  • AI Explainability UI: Visual system showing rationale (crew legality, adjacency restrictions, passenger connections).

  • Alerts & Exception Handling: Predictive focus zones with color-coded priorities; transparency in automatic optimizer actions.

  • Sandbox Mode: Safe planning environment to test changes without publishing, with confirmation layers.

  • Collaboration Dashboards: GroundOps and Network views, reducing phone-based coordination.

​

Algorithm Collaboration:

  • Worked with engineers on the VLSN optimization algorithm (which evaluates thousands of possibilities in seconds).

  • Translated constraints and objectives (crew legality, towing distance, adjacency penalties) into readable UI explanations.

  • Ensured outputs were displayed in 1–2 mins, interpretable under time pressure.

Impact

  • Improved trust & usability → planners adopted AI suggestions more readily once rationale was transparent.

  • Reduced stress during disruptions by shifting from reactive to predictive decision-making.

  • Laid groundwork for commercialization → design assets are now guiding adoption by other global airline partners.

  • Proved the value of UX in high-stakes, algorithm-heavy systems.

Key Takeaways

  • Tested interactive prototypes with Southwest planners.

  • Iterated on color coding for alerts until critical vs. low-priority conflicts were instantly distinguishable.

  • Refined sandbox mode after users asked for a safe testing space without impacting live ops.

  • Added explanation tooltips after feedback that “we need to know why the AI chose this gate.”

Get to know             better

GateAI is a confidential collaboration with Southwest Airlines. What you see here is a high-level overview.


I’d love to walk you through the full case study, including design artifacts and prototypes, in a one-on-one conversation. 

​

You can call or e-mail me directly, or simply fill out the form below to get in touch and know more.

bottom of page