Modeling practice and design principles in paleoinspired robotics: why deep time matters

Paleoinspired robotics is an emerging field at the intersection of paleontology, evolutionary biology, and robotics. It concerns both reconstructing extinct biomechanical systems and extracting design principles for novel robotic applications. This paper examines the methodological and epistemological implications of incorporating deep time into robotics, asking: What modeling practices underlie paleoinspired robotics? How does it differ from classical biorobotics and deep-time sciences? More broadly, why does incorporating deep time in the creation of physical (rather than abstract evolutionary) models matter? To address these questions, within paleoinspired robotics I distinguish paleo-robotics, which simulates past mechanisms based on deep-time data, from paleobionics, which selectively extract and repurpose extinct biological features for new technological applications. I examine how incorporating deep time influences the modeling practices within these subdisciplines compared to classical biorobotics and other historical sciences. I argue that while all these fields study complex systems, their engagement with temporality and the use of historical constraints in modeling differ significantly. While paleoinspired robotics shares methodological traits with biorobotics and other deep-time sciences, its modeling practices diverge due to the role of temporality. Deep time imposes constraints on paleo-robotics to reconstruct past mechanisms, aligning it with other phenomenon-driven and deep-time sciences. In contrast, paleobionics use deep time to selectively extract, decide what can be excluded, and repurpose extinct building blocks which, once integrated into robotic systems, can evolve in novel and unexpected directions rather than serving historical reconstruction. Hence, this paper clarifies the different modelling practices and common ground among these biologically inspired and deep-time oriented sciences.