Materials that make robots smart

Abstract

We posit that embodied artificial intelligence is not only a computational, but also a materials problem. While the importance of material and structural properties in the control loop are well understood, materials can take an active role during control by tight integration of sensors, actuators, computation, and communication. We envision such materials to abstract functionality, therefore making the construction of intelligent robots more straightforward and robust. For example, robots could be made of bones that measure load, muscles that move, skin that provides the robot with information about the kind and location of tactile sensations ranging from pressure to texture and damage, eyes that extract high-level information, and brain material that provides computation in a scalable manner. Such materials will not resemble any existing engineered materials, but rather the heterogeneous components out of which their natural counterparts are made. We describe the state-of-the-art in so-called “robotic materials,” their opportunities for revolutionizing applications ranging from manipulation to autonomous driving by describing two recent robotic materials, a smart skin and a smart tire in more depth, and conclude with open challenges that the robotics community needs to address in collaboration with allies, such as wireless sensor network researchers and polymer scientists.

Terrain-aware semantic mapping for cooperative subterranean exploration

“…Other geometric characteristics, like roughness (fitness to a plane), have been shown to improve navigation performance ( Ye, 2007 ). The incorporation of non-geometric features, such as friction ( Hughes et al, 2019 ; Rodríguez-Martínez et al, 2019 ) and collapsibility ( Haddeler et al, 2022 ), may also significantly improve navigation performance over adverse terrain, however, these features are often computationally expensive to calculate and less applicable to travel at low-moderate speeds. Learned terrain features ( Guan et al, 2022 ; Shaban et al, 2022 ; Seo et al, 2023 ) offer a cheap and fast approach for real-time traversability estimation, yet the efficacy of these methods remains dependent on costly training and poses challenges in adapting to completely unknown environments.…”

Three-dimensional printed cellulose nanofibers/carbon nanotubes/silicone rubber flexible strain sensor for wearable body monitoring

“…The rapid advancement of technology in the last two decades has led to the widespread utilization of flexible tactile sensors in various applications such as motion monitoring, [1][2][3][4] humancomputer interaction, 5,6 electronic skin, 7,8 intelligent robots, 9,10 and intelligent vehicles. 11 To ensure the necessary stretchability, different polymers are commonly used as flexible substrates, with popular choices including PU (polyurethane), 12,13 PDMS (polydimethylsiloxane), 14 and hydrogels.…”

Toward smart composites: Small-scale, untethered prediction and control for soft sensor/actuator systems

“…Devising a controller that uses in-material neural network predictions allows for composite materials to deeply embed sensing, computation and actuation, thereby leading to “materials that make robots smart”, 39 blurring the boundary between materials and computation. 40 This new class of materials are capable of making decisions based on their own proprioception without the need for communication with external computers or external observers and take advantage of existing electronics for advanced computation.…”