One view of materials development is to search for what materials correspond to empty spaces on a hypothetical multi-dimensional map of the properties of available materials. Following a biomimetic philosophy, for instance, it can be seen that tough ceramics and moldable short fiber composites with high moduli are possible but absent from the list of available synthetic materials. Likewise artificial muscle is missing, where the properties are defined as a developed stress of over 300 kPa, a linear contraction of 25 % and a response time of below one second. Currently dielectric elastomers come closest but have disadvantages [1,2].The actin-myosin muscle system provides the performance target for electroactive polymer actuators [3,4]. The process is driven chemically by the energy change from hydrolysis of the polyphosphate bond as ATP (adenosine triphosphate) binds to myosin and is converted to ADP (adenosine diphosphate) and phosphate. A simple chemical analogy suggests that muscle-like gels should be feasible but it is now clear that the task is much harder than it seems.Early work on gel actuation by Katchalsky-Katzir demonstrated that engines could be built using the chemical energy in diluting a lithium bromide solution to drive contraction and expansion of a collagen belt [5]. In essence, the collagen expands to take up the salt solution or contract to exclude the water. This change is both a molecular conformation change and a volume change. Other chemically driven gels, for instance polyacrylic acid fibers [6] which respond to a pH change, also rely on a solubility change giving rise to a volume change.