Case Study
MMO for EO
Making a complex automated planning system understandable and steerable by a single operator under pressure.
Client
Thales Alenia Space
Year
2025
Type
Enterprise UX
Role
UX Research, UI Design, Prototyping, User Testing
Deliverables
Interactive prototype, Design system extension, Backlog
Tools
Figma, FigJam, Notion
Brief
The design problem
The Multi-Mission Orchestrator coordinates acquisition requests across many satellites, sensors, and missions inside the ground segment. The orchestration logic already worked as a proof of concept, but it was not something a person could operate. An operator faces hundreds of requests owned by different mission teams, with acquisitions constantly refused or made unfeasible by orbital reality, and has to re-plan fast during time-critical events.
The first version surfaced the data without making it legible or steerable. My job was not to skin the algorithm. It was to make a powerful automated system one human could read, trust, and control.
Process
What I did
I ran the project as a UX-driven process. I co-defined requirements with stakeholders through workshops and interviews, then mapped the operator's journey end to end. The journey map became the shared reference for designers, developers, and stakeholders, and surfaced operator friction before it reached development rather than after.
From there I worked in design, test, learn cycles: prototypes from low to high fidelity, seven moderated usability sessions to watch where people hesitated or fought the system, and three full prototype iterations driven by what I observed. The work was structured against a backlog of eight epics and more than fifty features, and built on the Quantum design system to stay consistent with the ground-segment tools operators already use.
Decisions
The decisions that mattered
The redesign rests on three choices.
The first was to give the operator the whole picture at a glance. I organised the dashboard around operational state — from refused through planned, sensing, and downlinked — with a live timeline of upcoming satellite contacts underneath, so timing and status are always visible instead of inferred.
The second was to present many missions as one plan. Requests historically owned by separate teams are brought into a single workspace and orchestrated together. This breaks with the legacy ground-segment workflow, so the interface had to make a cross-mission plan feel trustworthy rather than alarming, pairing a unified request table with an interactive map and per-asset statistics.
The third, and the core of the work, was to keep the human in command of the automation. Rather than forcing a choice between a fully automatic plan and fully manual control, I designed two clean ways to correct the algorithm: override the orchestration criteria at the mission level and let the whole plan recompute, or intervene on a single request, promoting an earlier-executable acquisition while the rest of the plan re-adjusts around that one change. Automation does the heavy lifting; the operator keeps authorship and accountability.
The new timeline makes me feel like I'm ahead of the mission instead of catching up to it."
Reflections
Outcome
The work delivered a tested, stakeholder-validated V2 prototype that turns a working but unusable proof of concept into an operator-centered tool, grounded in a journey map and refined through seven usability sessions and three iterations.
Enterprise UX is an exercise in earning trust. Operators had been burned by redesigns before — tools that looked cleaner but made their job harder. The only way to build credibility was to show up at their desk and watch how they worked before drawing a single wireframe.