The Schedule

Dynamic locomotion in nature is inherently adaptable, allowing animals to modify their body structure and sensing capabilities according to environmental feedback. This adaptability is often enabled by modular muscular segments capable of contraction, extension, or combined movements. Inspired by this biological principle, modular robot design seeks to replicate such versatile locomotion by allowing reconfiguration and adaptation of their structural and functional elements. These bioinspired modular systems can emulate natural muscular movements, enhancing their capabilities through flexible and multi-functional configurations. Such modularity provides robots with the unique potential to exceed biological examples by adopting multiple bioinspired gaits within a single robotic platform. For instance, can one robot seamlessly transition between distinct crawling patterns exhibited by earthworms and snails? 

Plants may initially appear immobile and static, but they perform a wide range of complex functions that are based on highly dynamic processes. Composed of a few basic building blocks and consisting of a combination of dead and living tissues, plants are capable of adapting to environmental conditions over the long term (e.g., load-optimized growth), responding to short-term changes (such as self-repair, immune responses, and self-regulation), or even executing movements that can be faster than those of animals. A precise understanding of the interactions between material, structure, and function in plants opens the possibility of transferring this knowledge to technical problems and developing sustainable solutions (biomimetics). To analyze the functions of plant model organisms, we employ advanced medical imaging techniques to examine dynamic processes in plants in 3D over time (essentially in 4D). 

Many intelligent responses or movements of biological organisms are the outcome of seamless, integrative operations between various biological signals such as electric signals, chemical potentials and mechanical stimulations. In contrast, artificial intelligent systems built upon conventional electrical computing identify some inherent hurdles for computational operations between non-electrical signals. Here, we present a strategy for building an unconventional class of computational intelligence, termed “integrated mechano-intelligence”, which provides a seamless and integrative computational interface for autonomous soft machines. We show that the architected design and strategic assembly of soft bistable elements can form into a monolithic computational platform where nondispersive elastic solitary waves autonomously propagate through networked mechanical computing units. A systematic understanding of the computational propagation behavior of the solitary waves allows to establish a general design rule for integrated mechanical computing, and its effectiveness is verified both numerically and experimentally. The developed mechano-intelligence based on integrated mechanical computing realizes autonomous, intelligent soft machines in which the reception, transmission, computation and linkage of mechanical information with actuators are performed in a seamless, integrative manner. These findings would pave the way for future intelligent robots and machines that perform operations between non-electrical environmental agents.

Conventional robots are controlled by electronic feedback mechanisms that process input and regulate actuation via traditional electronic hardware. Most soft robots are incompatible with this type of hardware since their bodies are flexible and lightweight. Therefore, various other non-mechatronic methods are chosen to control soft robots. Soft robots made up of hydrogel components or polymers that can absorb solvents resemble living organisms in which water or solvents are used as a media for overall metabolic control. The efficient actuation and feedback needed for (autonomous) motion in these robots can be accomplished via physical/chemical ideas that take inspiration from biological processes such as osmosis and transpiration. This talk demonstrates a few examples of such simple yet effective ideas with some self-regulating plant robots and a solvent-regulated PDMS actuator. It is shown how complexity can be achieved by adding ‘a pinch of salt’ to a hydrogel system or through crystal sugar templating to manufacture PDMS sponges.

Nature fabricates lightweight materials displaying fascinating combinations of mechanical properties through benign, energy-efficient processes. These excellent mechanical properties are a result of an intricate interplay between structural control over the nm up to the mm length scales and locally varying compositions. We are far from reaching a structural and compositional control that can span a similar range of length scales. As a result of the inferior structural and compositional control, most manmade polymers possess mechanical properties that are inferior to those of natural counterparts. These limitations restrict their use for example for soft robotic applications. In this talk, I will introduce granular polymers including responsive hydrogels and elastomers that can be 3D printed into load-bearing granular materials possessing locally varying mechanical properties such that they deform in a pre-defined manner. I will show they can be functionalized to render them responsive to changes in the environment such as temperature, and humidity. By functionalizing the interstitial spaces with electrically conductive particles, we transform these materials into stretchable piezoresistive sensors. Moreover, I will demonstrate that the addition of an appropriate deep eutectic solvents transforms these granular materials into stretchable multi-modal sensors that can simultaneously detect and classify changes in temperature, humidity and stress. These sensors have the potential to render structures proprioseptic, a feature that might open up new opportunities in designing closed-loop robotic systems.