← Back to work

Senior Product Designer · 2023 · Percepto

An FAA rule was about to block Percepto’s US business. The fix was not a screen. It was getting four teams to see one system.

Clear the FAA block with one flow.

My role
Senior Product Designer
Focus
Remote ID
Outcome
US cleared

The stakes

A regulatory checkbox that was actually a revenue blocker.

In 2023, new FAA Remote ID rules required every drone flight in the US to broadcast operator and aircraft location. Percepto’s platform could not do it. Without compliance, there was no complete US offering. What looked like a regulatory checkbox was a revenue blocker.

What the data said

Pre-launch was already fragile. Remote ID would land on top of it.

Before touching Remote ID, I pulled the preflight failure reports. GPS issues, heading errors, and sensor faults were already blocking missions. I paired that data with field interviews to see where risk actually concentrated for operators. That changed the design target: not “add a compliance step,” but make the whole pre-launch sequence survivable.

Power BI preflight failure analysis showing technical failure frequency by category
Preflight failure reports: where risk concentrated before Remote ID entered the flow.

The key decision

Map the system before anyone drew a screen.

Remote ID crossed four surfaces with four different owners: the AIM desktop app, an SMS handoff, a mobile browser, and native OS permission prompts. The pressure was to start mocking screens and surface the dependencies as they appeared. I built one swimlane model of the entire flow first: every decision point, every failure path, every constraint. PM, Support, and Engineering argued against that one model, not four separate ones.

Remote ID cross-device swimlane showing decision points, error paths, and GPS constraints across AIM, SMS, and mobile
Swimlane detail: decision points, error paths, and constraints. Takeoff stayed off until the operator’s GPS was known.
End-to-end Remote ID narrative flow from mission start to compliance confirmation
End-to-end overview: a fast read for PMs and support leads who did not live inside AIM daily.

Three paths, one commitment

Stress-test the options before anyone builds.

An SMS link, dedicated URLs, and a desktop browser path. PM and Engineering tested each one before any build decision. The SMS handoff won. Operators did not need a new app, and it worked when the field connection was weak.

Solution and risk matrix comparing three Remote ID directions on value, viability, usability, and feasibility
Three directions stress-tested before implementation. Hidden constraints surface here, not in QA.

One mission flow across surfaces Percepto did not fully own

The operator starts a mission in AIM, receives an SMS on their phone, grants location through the browser, and sees confirmation back on the desktop. Recovery covers weak GPS, a denied permission, and a dropped connection. New components went into the AIM design system so the pattern could be reused.

  • AIM launch dialog: select a robot, set coordinates, and launch a security mission on the map
  • Mission initialization: the drone turns on, connects, and the checklist moves toward mission readiness

Launch a mission on the desktop. The next screen checks that the drone is on, connected, and ready.

  • Secondary control: password entry
  • Secondary control: connecting state
  • Pre-launch Remote ID check
  • Location permission prompt
  • Connection failure state with map

Mobile states across the SMS journey: auth, connect, pre-launch check, location prompt, and recovery.

AIM design system sheet for Remote ID on desktop and mobile
Remote ID patterns shipped into the AIM kit so the team could reuse location, SMS, and connection states.

Outcome

Surfaces
4
Requirement
FAA Remote ID
Market
US cleared

The flow shipped and met the FAA Remote ID requirement that unblocked Percepto’s US market.

What stayed with me

I put the location prompt, the SMS states, and the connection failure into the AIM design system. The next person could reuse them instead of drawing a new compliance step.

The flight still had to pass through a phone Percepto did not own. Takeoff stayed off until the operator’s location was known. The next designer has to be able to add a check without making that moment longer.