Ranjana Mehta, a professor of industrial and systems engineering, studies the intersection of human behavior, technology, and high-demand work environments; in particular, how fatigue influences performance when interacting with advanced systems, says this article on the University of Wisconsin-Madison website. She points out that robots and exoskeletons increasingly assist with dirty, dangerous, or demanding tasks, but they don’t replace human cognitive flexibility; instead, they shift how people execute their work.
In her research, she emphasizes that training, interface design, and system integration must anticipate human limitations. For example, her team is building an AI-driven VR training platform for police de-escalation scenarios that uses physiological data and eye-tracking to simulate realistic stress and fatigue conditions. On the fatigue front, Mehta highlights that fatigue doesn’t just slow physical movement; it affects decision-making, trust in automation, and the chance of over-reliance on systems. She recalls a field study during flooding where drone operators became so fatigued by day two that their performance equaled that of someone legally drunk.
She argues that when people are tired, they gravitate toward minimal-effort strategies: simplified decisions, less vigilance, greater reliance on automation, even when those tools aren’t perfect. This raises design challenges: systems must detect and accommodate fatigue, adapt interfaces accordingly, and maintain safety even when human attention is compromised.
Engineers and designers working on human-machine systems can’t assume optimal human performance, especially in high-stress or long-duration operations. Fatigue must be treated as a variable, not a nuisance, and human-centered design must incorporate real human states, not just ideal ones.