LEO ABREU / SELECTED WORK

PRODUCT DESIGN / FIELD OPERATIONS

PROPAGATE / KEYSQUARE LABS

Product Designer / 2025

Connect equipment without losing the task.

I redesigned an Android setup experience for field teams connecting radios, drones and other equipment. The goal was to make the next physical action and the connection result clear without removing the controls specialists rely on.

AT A GLANCE

40% less

Time on key tasks

25% higher

Reported task success

30% fewer

Clicks

Figures reported from client testing with the development team. Sample size, protocol and the calculation behind the success change were not documented.

01 / CONTEXT & ROLE

Technical capability was not the problem.

The project supported teams connecting equipment to an Android device, including workflows involving the TAK application. These users work with different hardware and configurations. The previous Plugnix Connect interface had technical depth, but it did not give someone an obvious route from opening the app to establishing a new connection.

The original screens put radio details, network information and VPN state close to the surface. At the same time, everyday setup competed with a long menu of routing, configuration and diagnostic options. For someone unfamiliar with the device or the app, seeing lots of available controls did not tell them which one mattered first.

I led the design work across discovery, stakeholder and user interviews, current and future flows, UI exploration, component documentation and delivery. I worked with the client and development team through testing and iteration. My focus was the usability of setup and feedback, not the design of the underlying network or radio protocols.

FIG. 01

Four screens from the previous Android app

Four screens from the previous Android app

Radio details, VPN states and the open configuration menu show why the next setup action was hard to identify.

02 / DISCOVERY & FLOW DEFINITION

First setup is not the same as adding another device.

I gathered product materials and spoke with people who understood how the application was used in different service areas. The work separated user actions, system information and decisions about configuration. A useful flow would show which device was being set up and whether the person was starting from scratch or returning to an existing setup.

The original discovery board maps first login for a single radio, a returning user adding an unmanned aircraft system (UAS), and a returning user adding other connections. Comparing these scenarios exposed the limits of a universal checklist: a hardware prerequisite or existing configuration changes the next step. The interface needed a familiar entry point and different instructions after that point.

The trade-off was to standardize the way users recognize progress without standardizing hardware procedures. I used the maps to define where to keep guidance consistent, where to show device-specific instructions and where diagnostics should remain available without becoming the default path.

01 / RECOGNIZE THE DEVICE

Select the equipment and connection scenario

02 / FOLLOW ITS INSTRUCTIONS

Show the next physical or app action

03 / CHECK THE RESULT

Distinguish progress from a problem

FIG. 02

Discovery: first login and returning-user paths

Discovery: first login and returning-user paths

Original FigJam excerpt. The flow compares first setup, adding UAS equipment and adding a new connection.

03 / VISUAL EXPLORATION

Test hierarchy before settling on a treatment.

I explored multiple interface versions, including light and dark treatments, before converging on patterns for primary actions, navigation and connection feedback. The point of the comparison was to check what the user notices first: the next required action, the current connection or the advanced controls.

I treated the controls and the status as separate kinds of information. A primary action says what to do next; status explains what the system is doing or has already done. The same visual priority should not be assigned to a warning, a normal technical detail and a call to continue.

The exploration file includes several home variants. Those screens show design work in progress; they are not claims that every theme, color or interaction was released or tested separately.

FIG. 03

Light, dark and action-hierarchy explorations

Light, dark and action-hierarchy explorations

Original design board with alternative home views. Read it as evidence of iteration, not as a gallery of live themes.

04 / GUIDED CONNECTION

Explain what the person does and what the radio does.

In the guided radio flow, the screen pairs an ordered instruction with an illustration of the physical connection. It tells users where to connect the cable and which configuration is handled automatically. That separation prevents the app from implying that someone needs to configure a setting that the device handles itself.

After the physical action, the interface must tell the person whether the connection is still in progress or whether they need to intervene. The case does not claim every error state shipped. It shows the design intent: guide the common path, make feedback visible, and keep diagnostics available when the normal path is insufficient.

One risk I would validate separately is interruption: after switching attention back to the equipment, can the person identify the last completed step and avoid repeating it? A clear screen in isolation is not enough if the task loses continuity between the app and the hardware.

FIG. 04

Radio connection instructions

Radio connection instructions

WHAT TO LOOK FOR

Original high-fidelity screen. The illustration supports the physical step, while the copy distinguishes user action from automatic configuration.

05 / DESIGN SYSTEM & HANDOFF

Give engineers the whole journey.

I organized the delivery file around device journeys instead of a flat inventory of screens. It connects onboarding to equipment-specific work and provides separate areas for home, settings, diagnostics and radio details. A developer can inspect how a state is reached rather than infer behavior from a single polished screen.

I built the component library, page templates and documentation alongside the flows. The button guidelines explain hierarchy, anatomy and states, so repeated actions can use the same patterns across hardware. This documentation matters when someone adds another path without the original designer in the room.

The handoff board shows the scope of those sequences, while the radio instruction earlier in the case shows the detail of one step. They answer different questions: what needed to be implemented and what the user would see at a specific point.

FIG. 05

Delivery map: device journeys and supporting screens

Delivery map: device journeys and supporting screens

Original handoff file with separate paths for radio types and unmanned aircraft, plus home, diagnostics, settings and radio details.

FIG. 06

Component guidance for engineers

Component guidance for engineers

Original button hierarchy and interaction states. Reusable specifications sit alongside the device flows.

06 / TESTING & REFLECTION

Separate test results from field performance.

The client and development team tested the work and reported 40% less time on key tasks, 25% higher task success and 30% fewer clicks. The original material does not record participant count, the exact protocol or whether the success change was relative or in percentage points. I treat these as reported test results, not operational telemetry or ROI.

The next validation would compare first-time setup and reconnecting known hardware, recording task completion, time, assistance and incorrect assumptions about connection state. I would include interruption and an unsuccessful connection so the evaluation covers recovery, not just the happy path.

My takeaway was to make a demanding task predictable without pretending that all devices behave the same. Standardize the cues, explain the specific action and retain access to technical detail when the user actually needs it.