Turning to Bionic Machine Intelligence | MIT News

Juncal Arbelaiz Mugica is Spanish and octopus is a common menu item in Spain. However, Arbelaiz appreciates octopuses and similar creatures in a different way, her research into the theory of soft robotics.

More than half of the octopus’s nerves are distributed across its eight arms, each of which has some degree of autonomy. Such distributed sensing and information processing systems have intrigued Arbelaiz, who is investigating how to design decentralized intelligence for artificial systems with embedded sensing and computing. At MIT, Arbelaiz, an applied mathematics student, is working on the fundamentals of optimal distributed control and estimation in her final weeks before completing her Ph.D. this fall.

Taking inspiration from the biological intelligence of invertebrates such as octopuses and jellyfish, her ultimate goal is to design novel control strategies for flexible “soft” robots that can be used in cramped or fragile environments, such as surgical tools or search and search. – Rescue missions.

“The softness of soft robots allows them to dynamically adapt to different environments. Think worms, snakes or jellyfish and compare their locomotion and adaptation to vertebrates,” Arbelaiz said. “It’s an interesting expression of embodied intelligence—the lack of a rigid skeleton provides advantages for some applications and helps deal more effectively with real-world uncertainty. But this extra softness also brings new systems theory challenges.”

In the biological world, the “controller” is often associated with the brain and central nervous system – it creates the movement instructions for the muscles to achieve movement. Jellyfish and some other soft creatures lack a centralized nerve center or brain. Inspired by this observation, she is now working on a theory that could use decentralized sensory information sharing to control soft robotic systems.

“Centralized intelligence can be difficult to achieve when sensing and actuation are distributed within the robot body and onboard computing power is limited,” she said. “Therefore, we need this kind of decentralized scheme that guarantees the desired global behavior despite only sharing sensory information locally. Some biological systems, such as jellyfish, are beautiful examples of decentralized control architectures—motion is It’s achieved without a (centralized) brain. It’s fascinating compared to what we can achieve with artificial machines.”

A smooth transition to MIT

Her graduate studies at the University of Navarra in San Sebastian led her to work on fluid dynamics with MIT professor John Bush. In 2015, he invited Arbelaiz as a visiting student to MIT to study droplet interactions.This led to their 2018 paper Physical Review Fluid, and her Ph.D. at MIT.

In 2018, her doctoral research was transferred to the Interdisciplinary Socio-Technical Systems Research Center (SSRC), now advised by Ali Jadbabaie, JR East Professor of Engineering and Chair of the Department of Civil and Environmental Engineering; and Anette “Peko” Hosoi, Associate Dean of the Faculty of Engineering, who is the Neil and Jane Pappalardo Professor of Mechanical Engineering and Professor of Applied Mathematics. Arbelaiz also frequently collaborates with Bassam Bamieh, associate director of the Center for Control, Dynamical Systems, and Computing at UC Santa Barbara. She says working with this advisory team has given her the freedom to explore the multidisciplinary research projects she has been working on for the past five years.

For example, she uses A systems-theoretic approach to design novel optimal controllers and estimators for systems with spatiotemporal dynamics and gain a fundamental understanding of the sensory feedback communication topologies required to optimally control these systems. For soft robotic applications, this amounts to ranking which sensory measurements are important to best trigger each “muscle” of that robot. Does robot performance degrade when each actuator only has access to the closest sensory measurements? Her research describes this trade-off between closed-loop performance, uncertainty, and complexity in spatially distributed systems.

“I am determined to bridge the gap between machine autonomy, systems theory and biological intelligence,” she said.

next chapter

The two-year Schmidt Science Fellowship funds postdoctoral research for young researchers in fields different from graduate work, which will allow Arbelaiz to further explore the intersection of biology and machine intelligence after graduation.

She plans to spend her postdoctoral time at Princeton with Professor Naomi Leonard and collaborate with researchers in systems biology, computer science, and robotics to explore the reliability and robustness of biological and artificial integration. Specifically, she is interested in understanding how biological systems can effectively adapt to different environments so that she can apply this knowledge to man-made systems, such as autonomous machines, whose susceptibility to noise and uncertainty creates safety concerns.

“I foresee an unprecedented revolution in autonomous and intelligent machines, thanks to the fruitful symbiotic relationship between systems theory, computation and (neuro)biology,” she said.

pay forward

Growing up in Spain, Arbelaiz was acutely aware of the privilege of getting a better education than her parents. Her father earned a degree in economics through independent study while working to support his family. His daughter inherited his persistence.

“The hardships my parents went through made them cherish self-study, lifelong learning and critical thinking,” she said. “They passed these values ​​on to me, so I grew up a curious and persevering person, passionate about science, and ready to seize every educational opportunity.”

To pass this on to others, she mentors STEM students who lack guidance or resources. “I firmly believe that we should promote talent everywhere, and mentoring can be a key driver in encouraging underrepresented minorities to pursue careers in STEM,” she said.

As an advocate for women in STEM, she is a member of the Executive Committee of MIT (GWAMIT) and MIT Graduate Women, and has participated in various panels and workshops. She also conducts field experiments for children, such as at the MIT Museum’s Girls’ Day event.

“As scientists, we have a responsibility to share our knowledge, inform the public about scientific discoveries and their implications, and raise awareness of the value of research and the need for investment.”

Arbelaiz has also supported MIT’s Covid-19 outreach efforts, including discussing mathematical models of the virus and translating her former mentor John Bush’s MIT Covid-19 indoor safety app into Basque.

This interest in putting her STEM knowledge into practice is due to her education at MIT.

“Massachusetts Institute of Technology has been one of the best experiences of my life so far: it has led to tremendous academic, professional and personal growth,” she said. “I share MIT’s taste for collaborative and multidisciplinary research, its appeal to intellectual challenges, and its passion for advancing science and technology for the benefit of humanity.”

Source link