This paper describes the implementation and validation of a fuzzy inference engine for estimating human affective state in real-time, using robot motions as the stimulus. The inference engine was tested with 36 subjects. To the authors' knowledge, this paper reports the first such trial that measures affective response to human-scale physical robot motions for a statistically significant population. The results demonstrate that affective state arousal can be detected using physiological signals and the inference engine. Comparison of results between the two planners shows that subjects report less anxiety and surprise with the safe planner.