We began with a focused usability audit for embedded devices. The review combined heuristic inspection of the interface with walkthroughs of real AV diagnostic tasks described by experienced technicians and engineers. The findings showed that previous designs had organised screens according to backend modules rather than technician workflows. Functions sat where they existed in the code, not where they were needed in the sequence of work.

A later redesign had concentrated on colours and icons but kept the same underlying structure. As a result, technicians still had to remember which mode contained which diagnostics, and they could lose context when switching between signal checks. The issue was not the visual style, but the lack of a coherent engineering tools UX model aligned with field practice.

The device runs on a small embedded display of 480 by 320 pixels with limited single point touch and modest processing resources. Touch targets had to remain generous enough for gloved use. Text had to remain legible at arm's length. The measurement device GUI therefore could not rely on gesture driven patterns or dense information layouts.

These constraints shaped concrete decisions. We limited menu depth and number of items per screen so that each view could show a full set of options without reducing font size below a comfortable threshold. We avoided animation and heavy graphical treatments to keep interaction crisp on the constrained hardware. Each screen was designed as a small, self contained state that technicians could interpret in a fraction of a second while also watching cables, displays and switchers.

We gathered requirements with the MSolutions product owner, the lead firmware engineer, and a group of senior AV technicians who use professional measurement software in daily work. Each group had distinct priorities. Engineers wanted full access to low level parameters. Technicians wanted fewer steps and clearer confirmation of results. Product management needed a structure that could support future features without another redesign.

We translated these inputs into a single strategy through tension-driven reasoning. For each screen of the professional instrumentation interface, we defined an explicit outcome: what decision the technician should be able to take at that moment. Existing functions were then reassigned to these outcomes, and conflicting priorities were resolved at strategy level rather than left to ad hoc decisions on individual screens. This created a stable basis for release planning and for later extension of the embedded device UX design.

The conceptual breakthrough came from treating the device as a guide through a standard AV diagnostic narrative rather than a collection of tools. The new model structures the measurement sequence as technicians actually experience it. One typical workflow, for example, starts with link integrity checks, continues with EDID and HDCP verification, then moves to resolution and color space validation on each display, and ends with a consolidated confirmation that the installation satisfies the defined profile.

In the new embedded GUI, each state points to the next logical action. Parameters only appear when they are needed for the current diagnostic step. The visual hierarchy emphasises one technical intention per screen and relegates secondary information to predictable positions. For technicians, the device now behaves like an experienced colleague that surfaces the right checks in the right order, rather than a drawer full of separate instruments.

We translated the new model into interactive prototypes and tested them with AV technicians who regularly work on multi monitor conference rooms and video walls. Sessions combined task based observation with short interviews. Technicians were asked to carry out realistic scenarios, such as identifying the cause of an incorrect resolution on one display in a signal chain that otherwise appears healthy.

Feedback focused on terminology, grouping of signal parameters, and the order in which results should appear when a fault is discovered. The core workflow did not require change, but many details were adjusted. Certain labels were revised to match the language technicians use on site. Some intermediate confirmation states were simplified to avoid hesitation. After these refinements one participant remarked that the embedded interface finally matched the way they already think when standing in front of a rack. The full loop of testing and iteration took two intensive days within the six week project window.

Once the embedded interaction model was stable, we extended the technical device UX to a laptop and mobile environment. The responsive architecture preserves the same sequence of technician workflows while using the larger surfaces to show relationships between measurements, historical values and reference profiles more clearly.

Technicians can now connect to the measurement device from a laptop during commissioning sessions or use a mobile interface during quick checks. The cross platform interface design allows remote control in tight spaces and better collaboration between colleagues on site and colleagues at a central operations centre. Because the conceptual model is identical, there is no need to learn separate behaviours for each platform.