Towards mechanically intelligent machines
Traditional, rigid-bodied robots are extensively employed in manufacturing, excelling in the efficient execution of specific programmed tasks. However, their adaptability is often limited. When engaging with unstructured environments autonomously, these robots face challenges, typically requiring an array of sensors and actuators. Nevertheless, recent advancements highlight the integration of physical intelligence into robot structures, enabling autonomous responses to environmental cues with fewer sensors, actuators, and controllers.
Here, we embrace large deformations, instabilities and multistability as paradigms to expand the functionality of robots. In the first part of this seminar, I will describe how instabilities can be exploited to generate complex and reprogrammable deformations out of uniform cylindrical shells. Shifting focus in the second part, I will discuss underactuated multistable linkages capable of executing reprogrammable sequences of motion with just a single input.
Katia Bertoldi is the William and Ami Kuan Danoff Professor of Applied Mechanics at Harvard University. Her research has been highlighted by many news sources including the BBC, and as of June 2020 had been cited over 11,000 times.
Bertoldi earned master’s degrees from the University of Trento in 2002 and from Chalmers University of Technology in 2003, majoring in Structural Engineering Mechanics. Upon earning a Ph.D. degree in Mechanics of Materials and Structures from University of Trento, in 2006, she joined the group of Mary Boyce at MIT as a post-doc. In 2008 Bertoldi became an Assistant Professor in Engineering Technology at University of Twente. In 2010 Bertoldi left the University of Twente to join the School of Engineering and Applied Sciences at Harvard University where she established the Bertoldi group focused on the study of the mechanics of materials and structures.
Bertoldi’s research contributes to the design of materials with a carefully designed meso-structure that leads to novel effective behavior at the macroscale. Bertoldi investigates both mechanical and acoustic properties of such structured materials, with a particular focus on harnessing instabilities and strong geometric non-linearities to generate new modes of functionality. Since the properties of the designed materials are primarily governed by the geometry of the structure (as opposed to constitutive ingredients at the material level), the principles Bertoldi discovers are universal and can be applied to systems over a wide range of length scales.