Energy in Robotics: An Interdisciplinary Challenge

Abstract

Advances in Robotics are challenging our definitions of what robots are and our perspective on how we interact with them. The conventional image of a robot as a large structure made of stiff, rigid links, is slowly being replaced by machines that are more lifelike. Now that we know robots are capable of outperforming humans in terms of strength or precision their role in manufacturing has expanded. As researchers in Robotics push the limits of what robots can be made of, at which scales, and what capabilities they offer, other segments of the economy are considering robots as technological solutions. Healthcare, telecommunications, even agriculture or shipping may become the next arenas for widespread deployment of robots. As robots are developed and deployed, based on different technologies and across a wide range of sizes, there is a growing need for a unified framework of discussion. In particular in the rapidly growing fields of soft and small robots, the importance of energy as a performance metric is paramount. How the robot stores energy, or harvests it from the environment sets the baseline for how much valuable work it can perform in the target environment. How the energy is converted from storage to mechanical work used in manipulation and locomotion, at which scales and with what efficiency are the key metrics for matching a robotic technology to a specific application. As robots are miniaturized, the macro scale distinctions between soft and rigid components become blurred, and energy metrics can be used to describe both technology types and allow for direct comparisons. In this new framework, multiple technologies can be discussed and evaluated across different robotic functionalities. For energy storage, the conventional electrical energy storage in a battery or supercapacitor, can be replaced with chemical fuels, storage in mechanical mechanisms, or even direct harvesting from the environment, for example through turboelectric nano generators. For actuators the broad range of technical solutions (e.g. fluidic soft actuators, chemo-mechanical transducers, dielectric elastomer actuators, piezoelectric actuators, etc.) can be compared across standard metrics, including specific energy and energy density, energy conversion efficiency, and bandwidth. Lastly, new components can provide multi functionality, for example actuators that offer opportunities for embedded sensing and unique control strategies. To showcase the recent advances in these rapidly growing fields, we have organized this special section of Advanced Intelligent Systems focusing on “Energy Storage and Delivery in Robotics”. The special section offers both original research work, as well as reviews and perspective articles from different disciplines, including haptics, mechanical, electrical, and chemical engineering, as well as materials science. The general themes and contributions are summarized below: At small scales and with powerful fabrication capabilities, new materials can be developed to push robots towards the performance limit of natural systems. Gracias et al. (article number 2000195) demonstrated solvent responsive self-folding of graphene architectures, an example of chemo-mechanical energy transduction. The researchers were able to show complex structures, including helices and pyramids deforming in response to chemical fuels, with no electronic components. In a similar venue, Ng et al. (2100085) built and operated self-sustained robots based on functionally graded elastomeric actuators carrying up to 22× their body weight. Borrowing lessons from biology, the researchers use soft actuators in combination with a more rigid exoskeleton to boost its performance. Aiming to provide a framework for discussions of energy in soft robotics, Stokes et al. (2000264) analyzed energy-based abstraction for soft robotic system development. In a more conventional robotic framework, Jusufi et al. (2000244) demonstrated modeling and control of a soft rob