Quantifying the effects of varying light-visibility and noise-sound levels in practical multimodal speech and visual gesture (mSVG) interaction with aerobots