I. Introduction
We envision that robust and reliable legged robots will support us not only at home and in our indoor work spaces. Ideally, they will work wherever humans go and beyond. Fields, forests, and mountains are examples of natural en-vironments to which vehicles have limited access. Where animals roam naturally, most wheeled and tracked vehicles are hindered by terrain roughness, slopes, available space, and natural substrates. Towards the goal of legged robots in natural environments, here we aim to design and characterize multi-segmented robot feet mounted to a bird-inspired robot leg. With animals as models in mind, we want to equip our robots with feet that mechanically adapt their shape to comply with rigid, uneven terrain and grab into soft terrain. Bipedal robots require feet with a range of viable center of pressure (CoP) points, or they must constantly and actively be balanced [1]. This criterion disqualifies com-monly mounted ball- and cylinder-shaped feet that otherwise perform well on quadruped legged robots [2], [3].