Skip to main content

Autonomous Soft Robots Without Electronics

Periodic Reporting for period 1 - ANSWER (Autonomous Soft Robots Without Electronics)

Reporting period: 2016-06-01 to 2017-08-31

Conventional robots usually consist of heavy rigid components, such as engines, gearboxes and rigid linkages that are made of high-density materials. Although they can perform complex movements and processes, they are typically not able to perform movements similar to those of biological models. Dielectric elastomer actuators (DEAs) allow flexible mechanisms to behave as artificial muscles. They typically consist of mechanically pre-strained elastomer membranes and compliant electrodes. They are lightweight and can produce impressive muscle-like strains. To control dielectric elastomer actuators, complex, expensive and external electronic control units are generally required, which often makes the practical application of DEA complicated and rather attractive of commercial products. However, dielectric elastomers can also act as sensors and piezoresistive switches (Dielectric elastomer switches - DESs), enabling the integration of monitoring and control functions in compliant components themselves. During the ANSWER project at the Biomimetics Laboratory at the University of Auckland and the TU Dresden, dielectric elastomer components are used in complex soft robotic systems. The aim of ANSWER is to integrate sensing, signal processing and actuation by the use of only flexible dielectric elastomer components in soft robotic structures without using conventional electronics. Based on the current knowledge of the DESs at the Biomimetics Laboratory, sensor-actuator systems comprising dielectric elastomer (DE) sensors, actuators and logic switches will be designed, to monitor, evaluate and react to certain environmental conditions. The developed laboratory scale processes will be transferred to modern production technologies at the Solid State Electronics Laboratory in Dresden in cooperation with the Werner-Hartmann-Zentrum for technologies of electronics.

Multi-functional sub units are distributed through the robotic structures (Figure 1). These integrated structures possess all functions required for self-diagnosis, -regulation and interaction with their environment. All these components only consist of polymer-conductive-filler mixtures. Figure 2 depicts an example signal flow chart of an autonomous, soft robot without conventional electronics. The whole robot can be considered as a compliant mechanical structure or combination of compliant mechanical structures, which can undergo significant deformation. Those deformations will be transferred to the dielectric elastomer membrane and, hence, to the DE-sensors. That causes a change in their capacitance C and resistance R. These changes are registered by the DE-logic unit, consisting of DESs, which again controls the charging and, thus, the actuation of dielectric elastomer actuators. These actuators then cause the reaction of the soft robot to the sensed environmental impulse. Thereby, the robotic structure is only supplied with a constant direct current. All necessary signal processing and generation of driving signals for actuators is done within the structure by soft dielectric components. Therefore, the ANSWER project not only pushes forward the development of soft robotics but also the development of flexible and, more importantly stretchable electronics.

The large amount of media coverage and the numerous activities to disseminate the project to non-scientific public, especially in New Zealand, supports the general visibility of engineering science in society. Soft robotics and biomimetics are widely known topics in New Zealand mainly because of the interesting projects of the Biomimetics Lab and their good communication via main stream media. The ANSWER project was able to support that communication, by demonstrating interesting and appealing results. We plan to keep up that good media coverage in the future, in order to back the acceptance and the understanding about the necessity of publically funded research.
In the first project phase, basic concepts of soft, biomimetc DE robot design have been developed. Functional sub-units, such as high-voltage logic, memory, signal generators, sensors and actuators have been developed and demonstrated. Mechanical models of the used elastomer materials have been derived analytically and numerically. A working finite element model was developed and validated experimentally. A first electronic free semi-soft robot (Trevor the Caterpillar) was designed, assembled and tested. A study for a dragonfly wing (Jule Dragonfly) was developed, assembled and tested.
Extensive testing and characterization of dielectric elastomer oscillators - so-called artificial central pattern generators (aCPGs) - has been done. A mathematical model of the electro-mechanical cub-components and the entire system was derived based on those measurements. During the secondment to FESTO studies for soft robots and their applications were conducted.
During the first project phase the dielectric elastomer technology for soft robotics was advanced. Several high voltage logic units have been developed and assembled soft signal processors and drivers for soft robots. First semi- and fully soft robots and robotic sub-components have been developed and showcased. Until the end of the project, we expect to be able to produce autonomous, soft robotic structures, using modern, scalable production technologies such as 3D- and ink-jet printing. The entire development of soft robotics and every advance in this field can be seen as beneficial to general society, since biologically inspired robots advance the acceptance of robotic systems in our everyday life and enable assistive medical devices and prosthetics of the next generation.
Figure 2: Signal flowchart of an autonomous soft robot without conventional electronic.
Figure 1: Schematic of biomimetic, robotic sturcture, possessing distributed multifunctional DEs.