Why HMI Design Matters More Than Most Engineers Think

A well-designed HMI is the difference between an operator who catches a developing problem in seconds and one who stares at a screen full of flashing colors while production grinds to a halt. In our experience building custom automation systems, the operator interface is frequently the last thing that gets attention during a project—and it shows. Poorly designed screens lead to slower response times, higher error rates, and frustrated operators who resort to workarounds that undermine the system you spent months engineering.

The ISA-101 standard provides a formal framework for HMI design, but you do not need to implement a full HMI philosophy document to make meaningful improvements. What follows are practical guidelines drawn from decades of building and commissioning automation equipment across industries ranging from automotive to medical device manufacturing.

Start With the Operator, Not the P&ID

The most common mistake in HMI design is treating the interface as a digitized version of the process and instrumentation diagram. Engineers naturally think in terms of equipment topology—pumps, valves, conveyors, and their physical relationships. Operators think in terms of tasks: What am I making? Is it running correctly? What do I need to do next?

Effective HMI design starts by understanding operator workflows. Spend time on the floor watching how operators interact with the current system. Identify what information they need at each stage of the process and what actions they take most frequently. Build the interface around those tasks, not around the physical layout of the equipment.

A task-based approach typically results in fewer screens, less navigation, and faster response to abnormal conditions. Instead of a screen for every piece of equipment, you end up with screens organized around operational states: startup, normal production, changeover, shutdown, and fault recovery.

Screen Layout and Information Hierarchy

Every HMI screen should answer three questions within the first two seconds of an operator looking at it: What is the current state? Is everything normal? Does anything need attention?

To achieve this, follow a consistent layout hierarchy across all screens:

Top zone — Status bar showing machine state (running, stopped, faulted), current recipe or product, and production counts. This area should be visible on every screen without exception.

Center zone — Primary process visualization. Show the current operation with only the information relevant to that context. Avoid the temptation to display every available data point. If an operator needs to see bearing temperature only during a specific phase, show it only during that phase.

Bottom zone — Navigation and command buttons. Keep navigation consistent across all screens. Operators should never have to hunt for the button that takes them back to the main screen.

Alarm banner — A dedicated area, typically at the top or bottom, that displays active alarms regardless of which screen the operator is viewing. This should never be obscured by pop-ups or navigation elements.

Color Standards That Actually Work

Color is one of the most abused elements in HMI design. Legacy systems often look like Christmas trees—every valve, motor, and sensor rendered in bright primary colors with animated elements competing for attention. When everything is colorful, nothing stands out.

The high-performance HMI approach, promoted by the ASM Consortium and ISA-101, uses a gray-scale base with color reserved for abnormal conditions. Here is a practical color scheme:

  • Gray tones — Normal operating conditions. Use varying shades of gray for equipment, piping, and backgrounds. This might feel bland at first, but operators quickly learn to scan for the absence of color as confirmation that everything is running correctly.
  • Green — Use sparingly. Green should indicate that a specific permissive or condition is met, not that equipment is simply running. When everything on the screen is green, operators become desensitized to it.
  • Yellow/Amber — Advisory conditions. Something has changed or is approaching a limit but does not yet require immediate action.
  • Red — Alarm conditions requiring operator action. Red should appear only when something genuinely needs attention. If your screens routinely show red during normal operation, your alarm configuration needs work.
  • Blue — Informational elements like setpoints, labels, and reference values.

Avoid using color as the sole indicator of state. Pair color changes with text labels, shape changes, or position changes. Roughly 8% of males have some form of color vision deficiency, and your operator pool likely includes someone affected.

Alarm Management Integration

Poor alarm management is the single biggest contributor to operator frustration and missed events. An HMI that generates hundreds of alarms per hour trains operators to ignore all of them. The goal is fewer than one alarm every ten minutes during normal operation—a target that feels aggressive until you audit how many of your current alarms are nuisance or chattering alarms that operators habitually acknowledge without reading.

Design your alarm displays with clear priority levels. Critical alarms should look and behave differently from advisory notifications. Include the following in every alarm message: what happened, where it happened, and what the operator should do about it. "High Temperature" is a bad alarm message. "Zone 3 barrel temperature 15°F above setpoint — check heater band and thermocouple" gives the operator actionable information.

Alarm shelving and suppression capabilities should be built into the HMI, but with appropriate security levels and automatic unshelving timers. No alarm should be permanently silenced without a management-of-change review.

Operators should be able to reach any screen from any other screen within two touches or clicks. A common and effective pattern is a three-level hierarchy:

  1. Overview screen — Shows the entire process at a high level with key performance indicators and status for each major section. This is the home screen and the screen that should appear after any period of inactivity.
  2. Area screens — One screen per major process section showing equipment detail, operating parameters, and local controls.
  3. Detail screens — Accessed from area screens for specific equipment diagnostics, trend displays, or configuration.

Avoid deep navigation trees. If operators need more than three touches to reach a critical control, the screen structure needs revision. Provide direct-access buttons for the most frequently used screens, and include a persistent "Home" button on every screen.

Real-time trend displays are among the most valuable tools you can provide to an experienced operator. An operator who has watched a temperature profile for months can spot a developing problem long before it triggers an alarm threshold.

Design trend displays with sensible defaults: appropriate time scales, auto-scaling Y-axes with meaningful ranges, and the ability to overlay related variables. Make it easy for operators to compare current behavior against historical runs or golden batches.

Avoid cluttering trend displays with too many variables. Four to six traces per trend window is a practical maximum. If operators need to see more, provide a way to toggle traces on and off rather than cramming everything onto one chart.

For production data, use contextual displays that show metrics relevant to the current operation. Tie this into your data acquisition and historian systems so operators can pull up historical context without leaving the HMI environment.

Security and Access Control

Not every operator needs access to every function. Implement role-based access control with at least three levels:

  • Operator — Can run the machine, acknowledge alarms, and view all screens. Cannot change setpoints or recipes.
  • Supervisor — Can modify setpoints within defined ranges, perform recipe changes, and access diagnostic screens.
  • Engineer/Maintenance — Full access including configuration, calibration, and alarm limit changes.

Auto-logout timers prevent screens from remaining in elevated access modes when a supervisor walks away. Keep the timeout short enough to be meaningful—five minutes is a reasonable default—but long enough that it does not interrupt active work.

Responsiveness and Performance

An HMI that lags behind the process is worse than no HMI at all. Screen refresh rates should be fast enough that operators perceive the display as real-time. For most process applications, a one-second update rate is adequate. For high-speed discrete applications like automated assembly systems, sub-second updates on critical indicators may be necessary.

Test your HMI performance under realistic conditions: full alarm loads, multiple screens open, historian logging active, and the network carrying normal traffic. Performance problems that appear only under load are the most damaging because they occur precisely when the operator needs the system most.

Practical Implementation Tips

When implementing these principles on a real project, prioritize the following:

  • Standardize screen templates before building individual screens. Define header layout, color palette, font sizes, button styles, and alarm banner placement once. Enforce consistency across the entire application.
  • Use a style guide that maintenance technicians and future integrators can follow. The HMI will be modified over the life of the machine, and without a guide, modifications gradually erode design consistency.
  • Conduct operator reviews at the 30%, 60%, and 90% completion points. Operators will identify usability problems that engineers miss. Build time into the project schedule for these reviews and the resulting changes.
  • Test with gloves. If your operators wear gloves, every touchscreen interaction must work with gloves. Button sizes, spacing, and touch targets all need to accommodate this constraint.

The Long-Term Payoff

A well-designed HMI reduces training time for new operators, decreases the frequency of operator errors, and improves response time to abnormal conditions. These are measurable outcomes. Track error rates, response times, and alarm acknowledgment patterns before and after an HMI redesign to quantify the improvement.

The investment in proper HMI design is modest compared to the cost of the automation system it controls, but the impact on daily operations is disproportionately large. An operator who trusts the interface and can find information quickly is an operator who keeps production running.

AMD Machines designs operator interfaces as an integral part of every automation system we build—not as an afterthought. Contact us to discuss how we approach HMI design for your application.