
In the basement of a decommissioned British power station, there's a mimic board the size of a living room wall. Built in the early 1950s, it uses backlit tiles, mechanical indicators, and hand-painted diagrams to represent the state of an entire electrical distribution network in real time.
It's beautiful. It's also, in many ways, better than what replaced it.
The Problem with Modern HMIs
Walk into any steel plant control room built after 1995 and you'll find a bank of monitors running SCADA screens — grey backgrounds, dense numeric displays, alarm lists scrolling faster than anyone can read. The information density is extraordinary, but the legibility is poor.
Operators cope by ignoring most of the display. Studies of control room behavior consistently find that operators monitor 5–8 key variables and rely on alarms for everything else. The remaining screen real estate is, functionally, wallpaper.
The mimic board solved this differently. Its designers understood something that modern HMI vendors have largely forgotten: the primary job of a process display is not to show data. It's to make abnormal conditions visible without requiring the operator to look for them.
What the Mimic Board Got Right
Three design principles made the mimic board effective:
1. Spatial consistency. The board was a physical map of the system. Generators were at the top, load centers at the bottom, transmission lines ran left to right. An operator who'd worked with the board for a month could locate any component without thinking. Modern HMI navigation — nested menus, tabbed displays, pop-up windows — destroys spatial memory.
2. Change detection over absolute values. The board used colored indicators that changed state: green for normal, red for tripped, amber for manual override. The operator didn't need to read a number and compare it to a setpoint. The board did the comparison and showed only the result.
3. Information hierarchy through physical scale. Important systems got more wall space. Auxiliary systems were smaller. The visual hierarchy matched the operational hierarchy — not because a UX designer decided it should, but because the engineers who specified the board were also the people who operated the plant.
Implications for Industrial Software
At PRAX, we think about these principles constantly. When we build operator-facing tools, the question is never "what data can we show?" but "what decisions does this person need to make, and what's the minimum information that supports them?"
This leads to interfaces that look sparse by SCADA standards. Large type. Few numbers. Aggressive use of color to encode state. And spatial layouts that map to the physical plant, not to the database schema.
The mimic board builders understood something we keep having to relearn: in high-stakes, time-pressured environments, clarity beats completeness every time.