Abstract:

In my talk, I will address a prevalent challenge as we witness the increasing ubiquity of machines with artificial intelligence.

How do we design and validate human-machine systems that are tailored for the cognitive abilities of their human users? To date, theusability of human-machine systems is evaluated in terms of either performance (e.g., task completion times) or subjective recall (e.g., interviews). Unfortunately, these measurements are an indirect consequence of the users’ ability to use a system.

They do not directly reflect a user’s level of arousal, frustration, engagement, and more. I will provide examples from my research---in the domains of surface transport, aviation, augmented-reality, etc---to demonstrate how user states during human-machine-interactions can be implicitly measured and inferred with the use of gaze-tracking and electroencephalography. More importantly, my talk will address how this research informs the design of human-machine systems that can adapt to the needs and cognitive abilities of their users.