About

Eva Adnan Kanso is a professor and the Z.H. Kaprielian Fellow in Aerospace and Mechanical Engineering at the University of Southern California. She joined USC in 2005 and has held a variety of academic and research positions, including a two-year postdoctoral position in Computing and Mathematical Sciences at Caltech. Her educational background includes a Ph.D. in Mechanical Engineering, an M.S. in Mechanical Engineering, and an M.A. in Mathematics from the University of California at Berkeley, as well as a Bachelor of Engineering degree from the American University of Beirut with distinction. Kanso's research interests focus on fundamental problems in the biophysics of cellular and subcellular processes and the physics of animal behavior, both at the individual and collection levels. A central theme in her work is the role of the mechanical environment, specifically the fluid medium and fluid-structure interactions, in shaping and driving biological functions. She founded the Bioinspired Motion Lab at USC, where she leads a team of students and post-docs working on bio-inspired aerial and aquatic locomotion, with applications to autonomous soft robotic vehicles, engineering design, and collective behavior. Her research also includes biophysical modeling of bacteria and ciliary systems, with applications to biomedical flows, active materials, and microfluidic manipulation and transport. Kanso has held visiting positions at prestigious institutions such as Princeton University, the Laboratoire LadHyX at the Ecole Polytechnique, the Courant Institute of Mathematical Sciences, the Simons Foundation, and the Ecole Supérieure de Physique et de Chimie Industrielles. Her contributions have been recognized through awards including the NSF CAREER Award, the USC Mentoring Award, and the American University of Beirut Distinguished Young Alumnus Award.

Research topics

• Physics
• Artificial Intelligence
• Computer Science
• Biology
• Quantum mechanics
• Biological system
• Management science
• Statistical physics
• Classical mechanics
• Engineering
• Biophysics
• Data science
• Cell biology
• Materials science
• Genetics

Selected publications

• Optimal feeding in swimming and attached ciliates

  Journal of Fluid Mechanics · 2025-01-20 · 4 citations

  articleOpen accessSenior authorCorresponding

  Ciliated microorganisms near the base of the aquatic food chain either swim to encounter prey or attach at a substrate and generate feeding currents to capture passing particles. Here, we represent attached and swimming ciliates using a popular spherical model in viscous fluid with slip surface velocity that affords analytical expressions of ciliary flows. We solve an advection–diffusion equation for the concentration of dissolved nutrients, where the Péclet number ( $Pe$ ) reflects the ratio of diffusive to advective time scales. For a fixed hydrodynamic power expenditure, we ask what ciliary surface velocities maximize nutrient flux at the microorganism's surface. We find that surface motions that optimize feeding depend on $Pe$ . For freely swimming microorganisms at finite $Pe$ , it is optimal to swim by employing a ‘treadmill’ surface motion, but in the limit of large $Pe$ , there is no difference between this treadmill solution and a symmetric dipolar surface velocity that keeps the organism stationary. For attached microorganisms, the treadmill solution is optimal for feeding at $Pe$ below a critical value, but at larger $Pe$ values, the dipolar surface motion is optimal. We verified these results in open-loop numerical simulations and asymptotic analysis, and using an adjoint-based optimization method. Our findings challenge existing claims that optimal feeding is optimal swimming across all Péclet numbers, and provide new insights into the prevalence of both attached and swimming solutions in oceanic microorganisms.

  PublisherOA PDFDOI

• Self-reorganization and Information Transfer in Massive Schools of Fish.

  PubMed · 2025-06-03

  preprintOpen accessSenior author

  The remarkable cohesion and coordination observed in moving animal groups and their collective responsiveness to threats are thought to be mediated by scale-free correlations, where changes in the behavior of one animal influence others in the group, regardless of the distance between them. But are these features independent of group size? Here, we investigate group cohesiveness and collective responsiveness in computational models of massive schools of fish of up to 50,000 individuals. We show that as the number of swimmers increases, flow interactions destabilize the school, creating clusters that constantly fragment, disperse, and regroup, similar to their biological counterparts. We calculate the spatial correlation and speed of information propagation in these dynamic clusters. Spatial correlations in cohesive and polarized clusters are indeed scale free, much like in natural animal groups, but fragmentation events are preceded by a decrease in correlation length, thus diminishing the group's collective responsiveness, leaving it more vulnerable to predation events. Importantly, in groups undergoing collective turns, the information about the change in direction propagates linearly in time among group members, thanks to the non-reciprocal nature of the visual interactions between individuals. Merging speeds up the transfer of information within each cluster by several fold, while fragmentation slows it down. Our findings suggest that flow interactions may have played an important role in group size regulation, behavioral adaptations, and dispersion in living animal groups.

  PublisherOA PDF

• Hierarchical equivariant graph neural networks for forecasting collective motion in vortex clusters and microswimmers

  Communications Physics · 2025-11-20

  articleOpen access

  Abstract Data-driven modeling of collective dynamics is a challenging problem because emergent phenomena in multi-agent systems are often shaped by short- and long-range interactions among individuals. For example, in bird flocks and fish schools, flow coupling plays a crucial role in emergent collective behavior. Such collective motion can be modeled using graph neural networks (GNNs), but GNNs struggle when graphs become large and often fail to capture long-range interactions. Here, we construct hierarchical and equivariant GNNs, and show that these GNNs accurately predict local and global behavior in systems with collective motion. As representative examples, we apply this approach to simulations of clusters of point vortices and populations of microswimmers. In these systems, our approach is more accurate and faster than a fully-connected GNN. Specifically, only our approach conserves the Hamiltonian for the point vortices and only our approach predicts the transition from aggregation to swirling for the microswimmers.

  PublisherOA PDFDOI

• Tube feet dynamics drive adaptation in sea star locomotion

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-04-24 · 2 citations

  preprintOpen access

  Abstract Sea stars use hundreds of tube feet on their oral surface to crawl, climb, and navigate complex environments—all without the coordination of a central brain. While the morphology of tube feet and their role as muscular hydrostats are well described, the dynamics underlying their locomotion remain poorly understood. To investigate these dynamics, we employed an optical imaging method based on frustrated total internal reflection to visualize and quantify tube foot adhesion in real time across individuals of Asterias rubens spanning a wide size range. Our results reveal an inverse relationship between crawling speed and the duration of tube foot contact with the substrate. This suggests that sea stars regulate locomotion by modulating foot-substrate interaction time in response to body load. To test this, we conducted perturbation experiments using 3D-printed backpacks that increased body mass by 25% and 50%, along with numerical simulations based on a mechanistic model incorporating decentralized feedback control of the tube feet. The added load significantly increased adhesion time, supporting the role of a load-dependent mechanical adaptation. We further investigated inverted locomotion, both experimentally and through simulation, and found that tube feet adjust their contact behavior when the animal is oriented upside down relative to gravity. Together, our findings demonstrate that sea stars adapt their locomotion to changing mechanical demands by modulating tube foot-substrate interactions, revealing a robust, decentralized strategy for navigating diverse and challenging terrains.

  PublisherOA PDFDOI

• Cooperative hydrodynamics accompany multicellular-like colonial organization in the unicellular ciliate Stentor

  Nature Physics · 2025-03-31 · 14 citations

  articleOpen access
  PublisherOA PDFDOI

• Sensing flow gradients is necessary for learning autonomous underwater navigation

  Nature Communications · 2025-03-28 · 11 citations

  articleOpen accessSenior authorCorresponding

  Aquatic animals are much better at underwater navigation than robotic vehicles. Robots face major challenges in deep water because of their limited access to global positioning signals and flow maps. These limitations, and the changing nature of water currents, support the use of reinforcement learning approaches, where the navigator learns through trial-and-error interactions with the flow environment. But is it feasible to learn underwater navigation in the agent’s Umwelt, without any land references? Here, we tasked an artificial swimmer with learning to reach a specific destination in unsteady flows by relying solely on egocentric observations, collected through on-board flow sensors in the agent’s body frame, with no reference to a geocentric inertial frame. We found that while sensing local flow velocities is sufficient for geocentric navigation, successful egocentric navigation requires additional information of local flow gradients. Importantly, egocentric navigation strategies obey rotational symmetry and are more robust in unfamiliar conditions and flows not experienced during training. Our work expands underwater robot-centric learning, helps explain why aquatic organisms have arrays of flow sensors that detect gradients, and provides physics-based guidelines for transfer learning of learned policies to unfamiliar and diverse flow environments. Aquatic animals outperform robotic vehicles in underwater navigation due to robots’ limited access to GPS and flow maps in deep water. The authors report that to successfully learn navigation, an agent must sense both local flows and flow gradients, enabling adaptable and robust policies under unfamiliar conditions.

  PublisherOA PDFDOI

• Flow physics of nutrient transport drives functional design of ciliates

  Nature Communications · 2025-05-04 · 6 citations

  articleOpen accessSenior author

  Phagotrophy, the ability of cells to ingest organic particles, marked a pivotal milestone in evolution, enabling the emergence of single-celled eukaryotes that consume other organisms and leading to multicellular life. However, reliance on food particles also created a mechanical challenge—how to coordinate the transfer of particles from the exterior environment to the cell interior? Here, we investigate this important link using mechanistic models of ciliates, a clade of single-celled eukaryotes that either swim or attach and generate feeding currents to capture prey. We demonstrate that ciliates optimize their feeding efficiency by designating a specific portion of the cell surface as a ‘mouth,’ and optimal cilia coverage varies by life strategy: for sessile ciliates, prey encounter is most efficient when cilia are arranged in bands around oral structures while ciliates that swim display diverse ciliary arrangements that meet the cell’s nutritional needs. Importantly, beyond a threshold of doubling nutrient uptake, further increases in feeding flux do not seem to be a dominant selective force in cell design. The evolution of phagotrophy by microbes required effective particle transport and ingestion, enabling the rise of ciliates as key grazers in aquatic ecosystems. This study shows that the morphological adaptations of ciliates for phagotrophy were shaped by hydrodynamic forces.

  PublisherOA PDFDOI

• Nutrient transport in concentration gradients

  Physical Review Fluids · 2025-09-24

  articleSenior author
  PublisherDOI

• Traveling waves in a continuum model for schooling swimmers

  ArXiv.org · 2025-07-08

  preprintOpen access

  The complex formations exhibited by schooling fish have long been the object of fascination for biologists and physicists. However, the physical and sensory mechanisms leading to organized collective behavior remain elusive. On the physical side in particular, it is unknown how the flows generated by individual fish influence the collective patterns that emerge in large schools. To address this question, we here present a continuum theory for a school of swimmers in an inline formation. The swimmers are modeled as flapping wings that interact through temporally nonlocal hydrodynamic forces, as arise when one swimmer moves through the lingering vortex wakes shed by others, leading to a system of time-delay-differential equations. Through coarse-graining, we derive a system of partial differential equations for the evolution of swimmer density and collective vorticity-induced hydrodynamic force. Linear stability analysis of the governing equations shows that there is a range of swimmer densities for which the uniform (constant-density) state is unstable to perturbations. Numerical simulations reveal families of stable traveling wave solutions, where a uniform school destabilizes into a collection of densely populated "sub-schools" separated by relatively sparse regions that move as a propagating wave. We find that distinct propagating waves may be stable for the same set of kinematic parameters. Generally, our results show that temporally nonlocal hydrodynamic interactions can lead to rich collective behavior in schools of swimmers.

  PublisherOA PDFDOI

• Noise-Induced Collective Memory in Schooling Fish.

  PubMed · 2025-07-21

  preprintOpen accessSenior author

  Schooling fish often self-organize into a variety of collective patterns, from polarized schooling to rotational milling. Mathematical models support the emergence of these large-scale patterns from local decentralized interactions, in the absence of individual memory and group leadership. In a popular model where individual fish interact locally following rules of avoidance, alignment, and attraction, the group exhibits collective memory: changes in individual behavior lead to emergent patterns that depend on the group's past configurations. However, the mechanisms driving this collective memory remain obscure. Here, we combine numerical simulations with tools from bifurcation theory to uncover that the transition from milling to schooling in this model is driven by a noisy transcritical bifurcation where the two collective states intersect and exchange stability. We further show that key features of the group dynamics - the bifurcation character, transient milling, and collective memory - can be captured by a phenomenological model of the group polarization. Our findings demonstrate that collective memory arises from a noisy bifurcation rather than from structural bistability, thus resolving a long-standing ambiguity about its origins and contributing fundamental understanding to collective phase transitions in a prevalent model of fish schooling.

  PublisherOA PDF

Recent grants

• Collaborative Research: Geometric Time Integrators for Mechanical Dynamical Systems

  NSF · $125k · 2008–2012

• INSPIRE: How Do Cilia-driven Flows Shape the Chemical and Mechanical Environment in Bacteria-Host Associations?

  NSF · $1.0M · 2016–2020

• CAREER: Modeling and Control of Solid-Fluid Interactions in Aquatic Locomotion

  NSF · $400k · 2007–2013

• Collaborative Research: CPA-G&V-T: Aquatic Propulsion Laboratory

  NSF · $317k · 2008–2013

• UNS: Collaborative Research: Crossing the boundary: motion of solid objects across air-liquid interfaces

  NSF · $222k · 2015–2019

Frequent coauthors

• Janna Nawroth

  113 shared

• Laura von Schledorn

  84 shared

• Ruth Olmer

  Southern California University for Professional Studies

  84 shared

• Annemarie van Schadewijk

  University of Southern California

  83 shared

• Ayse T. Sahin

  Pioneer (United States)

  81 shared

• Feng Ling

  66 shared

• Sarah Glasl

  65 shared

• John H. Costello

  Providence College

  56 shared

Education

• Ph.D., Mechanical Engineering

  University of California, Berkeley

  2003

• M.A., Mathematics

  University of California, Berkeley

  2002

• M.S., Mechanical Engineering

  University of California, Berkeley

  1999

Awards & honors

• USC USC Mentoring Award, faculty mentoring graduate students…
• American University of Beirut Distinguished Young Alumnus Aw…
• National Science Foundation NSF CAREER Award (2006)
• University of California, Berkeley Outstanding Graduate Stud…
• Lebanese-American Association (LAA) Lebanese-American Associ…