--------------------------------------------
i n v e s t i g a t o r s :
Keith Evan Green, PI

--------------------------------------------
f u n d i n g :

(NSF pending)

--------------------------------------------
r e s e a r c h  q u e s t i o n :
Undertaking a “wicked” problem, can a human and a robot, collaborating as team partners, exhibit more productive, creative thinking than a human working alone?


COMPREHEND

The challenge: In recent years, robots working in close proximity to humans have become a reality. Categorically, NSF's NRI solicitation categorizes these robots as surgical, sentry, and household robots that “work beside, or cooperatively with, people to extend or augment human capacities." By definition then, all of these robots are working in close physical proximity to humans, some of them are working in cooperation with humans, but few of them are contributing to the human-machine partnership higher-order, thinking-and-making activities highly suited for addressing the kinds of “wicked” problems of society.

Out interdisciplinary team proposes to investigate intimate, human-environment cooperation that harnesses the best capabilities of people, machines, and their physical surroundings. We argue that, working together, this partnership—human being and robotic room—can advance the partners' performance well beyond the capabilities of either one performing alone.

The objective of this research is to advance the promise of human-robot cooperation in, specifically, a robotic physical environment that partners with human inhabitants to address complex problems in ways machines and humans alone cannot. The robot, COMPREHEND, takes its name as the acronym for COMpressed-Pattern, Robotic and Habitable, ENvironmental Design. Our team spanning architecture, robotics, and psychology will: (1) design COMPREHEND, a physically reconfigurable, architectural-robotic module occupying the space of a car interior and the smallest habitable room permitted by building code; and (2) evaluate, for creativity, the performance of this robot partnered with humans.

We are testing two hypotheses:

Hypothesis 1: The use of COMPREHEND in a human-robot partnership will result in a greater number of creative ideas being produced for a given wicked, thinking-and-making challenge as compared to (a.) the use of only the screen-based software offered by COMPREHEND, and (b.) the use of no software support (i.e. pen and paper).

Hypothesis 2: The use of COMPREHEND in a human-robot partnership will result in more creative (i.e. both novel and purposeful) ideas being produced for a given wicked, thinking-and-making challenge as compared to (a.) the use of only the screen-based software offered by COMPREHEND, and (b.) the use of no software support (i.e. pen and paper).

To test these two hypotheses, we propose prototyping COMPREHEND by way of an iterative design and evaluation process, and then testing a human-COMPREHEND partnership in lab experiments

Intellectually, the proposed research expects to contribute: a fully functioning prototype representing a design exemplar of a compressed-pattern, architectural robotic environment; a novel, robot-interaction modality based on rigid and/or compliant robots; an understanding of the diverse needs, challenges, and opportunities humans encounter when partnering with a robotic environment on complex problems; and, foremost, a pattern language elaborating an ecology of creativity for human-robot cooperation. The investigators emphasize that COMPREHEND, while drawing on a rich body of prior accomplishments, should be recognized as an early effort with great promise for human-machine interaction, particularly given recent theorizing on post-humanism (e.g., Hayles) in which human-machine collaboration exceeds the capabilities of humans.