An AI-assisted screening tool that reviews industrial HMI and SCADA screenshots for cognitive-load and human-factors concerns.


Years of changes can add colors, indicators, and alarms without a consistent design system.
Too many competing signals can hide the condition that needs an operator’s attention.
Specialist reviews take time, so large screen inventories may go years between checks.
Vantage combines a vision-language model with explicit checks based on ISA-101 and NUREG-0700 guidance. It highlights risky regions and explains where a qualified reviewer should look first.
Submit a static HMI or SCADA screenshot for analysis.
Extract layout features and check color, density, navigation, alarms, and contrast.
Generate a risk score, heatmap, and structured findings for review.
Flags decorative or excessive color that can weaken alarm meaning.
Reviews how much information competes within a single display.
Identifies controls and pathways that may slow operator scanning.
Surfaces text and visual elements that may be hard to read.
The project analyzes static screenshots and has been validated on a limited sample. Its output can vary and does not replace a certified human-factors audit.
Planned work includes live SCADA integration, scheduled monitoring, secure on-device models, and coverage across full screen inventories.