Final Report Summary - PIPA (Paleo-robotics and the Innovations of Propulsion in Amphibians)
A vertebrate limb is an exquisite mechanical device; it can accomplish many varied tasks from propelling a jump to accurately reaching for a target. Among vertebrate animals, frogs are famous for the great power, speed and control of their legs as they engage in various behaviours from jumping to running to swimming to climbing to digging. Our ERC project used frogs as a model to better understand how muscles and bones work together to enable a limb to be multi-functional. The question we asked was: what are the key anatomical and physiological features of a frog limb which enable such astounding multi-functionality? A great challenge in this research was the lack of laboratory techniques to measure the relationship between limb function and musculoskeletal anatomy. To tackle this, we developed several new methodologies and devices enabling us to computationally manipulate musculoskeletal anatomy to observe the outcome on animal behaviour. One such methodology we call "Reverse Haptics" which is a motorised robotic frog foot controlled by a real-time simulation of a frog (inside a PC). With this tool we could manipulate musculoskeletal features in software to observe changes in the behaviour of the robotic hardware. Other methods we developed are purely computational simulations (based on non-invasive observation) to greatly reduce the need for manipulating live animals. Our main scientific findings give evidence that 1) frog limbs are astonishingly versatile, not because of any specific unique anatomical features, but due to the complex 3D motions enabled by their leg joints and, 2) frog muscles (and muscles of most vertebrates) posses properties that allow the limb to autonomously adapt to different conditions (e.g. different terrains) and, 3) frog limbs (and perhaps other vertebrates) may move according to a fundamental rule governing how their joints are coordinated. Our scientific insights have advanced our understanding of animal limbs, but also hope to stimulate novel ways of approaching problems in prosthetic limb design. Moreover, the techniques developed will hopefully be used towards research applications in muscular ageing and disease.