Dielectric elastomer transducers are in essence deformable elastic capacitors. Coating an electrically insulating elastomer foil on both sides with stretchable electrodes forms such a stretchable capacitor. Dielectric elastomer transducers are capable of converting electrical into mechanical energy and vice versa. Applying a voltage to the electrodes deforms the soft dielectric elastomer - the transducer operates as actuator. At the reverse mode operation, charges are applied to the electrodes and external forces deform the elastomer foil. The elastomer deformation changes the electrical energy of the charges. Being based on soft polymeric materials assembled in a simple arrangement, dielectric elastomer transducers feature a high specific energy density, are light-weight and potentially low-cost. I combined the polymeric dielectric elastomer with soft ionically conductive hydrogel electrodes via a novel cyanoacrylate-based instant bonding technique. The low Young's modulus, the excellent transparency and the high stretchability of the hydrogel electrodes facilitate new designs for dielectric elastomer based adaptive optic elements. Here I report on a novel focus tunable lens with instantly bonded ionic hydrogel electrodes. The high optical transparency of both the hydrogel and the bonding interface enable placing the electrodes in the optical path. Our adaptive lens is capable of changing the focal length by 110% and operating at frequencies up to 10 Hz. Due to their potential small size and short response time, such variable optical systems based on transparent deformable materials are in demand where space is narrow, e.g., in mobile phones and endoscopes. Beyond the development of focus adaptive lenses, the instantly bonded hydrogel electrodes developed during my doctoral research study are readily applicable for dielectric elastomer generators (DEGs). Our proof-of-concept DEG with soft hydrogel electrodes converts 497 mJ mechanical input energy into 54 mJ electrical output energy per cycle. Small-scale energy harvesting devices are desired as power sources for portable consumer electronics like mobile phones, or untethered sensor nodes, whereas large-scale energy harvesting systems are promising for ocean wave energy harvesting. DEGs typically require an input bias voltage for their energy conversion process. I developed a DEG-inspired design for an electrostatic energy converter (EEC) capable of operating without bias voltage supply. The electret-biased EEC, being designed to work at low frequencies with little mechanical input energy, facilitates a remarkably high mechanical-to-electrical energy conversion efficiency of about 60%. In addition, this thesis discusses my work on developing a novel hybrid, low-cost simulator for enhancing the training of epidural needle insertion procedures. The patient phantom of the hybrid simulator combines realistic force feedback imitating the haptics of soft biological tissue with electronic sensing of the needle tip position via soft deformable electrodes. The real-time position information is expected to serve as guide for efficient training.