Efficient navigation in swarms often relies on the emergence of decentralized approaches that minimize traversal time or energy. Stigmergy, where agents modify a shared environment that then modifies their behavior, is a classic mechanism that can encode this strategy. We develop a theoretical framework for stigmergic transport by casting it as a stochastic optimal control problem: agents (collectively) lay and (individually) follow trails while minimizing expected traversal time. Simulations and analysis reveal two emergent behaviors: path straightening in homogeneous environments and path refraction at material interfaces, both consistent with experimental observations of insect trails. While reminiscent of Fermat's principle, our results show how local, noisy agent+field interactions can give rise to geodesic trajectories in heterogeneous environments, without centralized coordination or global knowledge, relying instead on an embodied slow fast dynamical mechanism.

Patterns in non-equilibrium systems—from phase separation to tissue morphogenesis—arise from microscopic stochastic dynamics but are most easily specified and measured at macroscopic scales. This work introduces a unified physics-based framework for generative control of such patterns that bridges particle (Lagrangian) and field (Eulerian) descriptions. Starting from overdamped Langevin dynamics for interacting particles with internal states, this work defines smoothed mesoscopic fields and a free-energy functional that governs macroscopic evolution. The resulting flows are projected back to particles via smoothed-particle hydrodynamics, closing a micro-macro modeling loop. Control objectives are posed at the field level but actuated at the particle level, enabling scale-dependent design. To solve the resulting high-dimensional stochastic control problem, an adjoint-based path integral control (APIC) is introduced, which marries an exact Feynman-Kac linearization of the Hamilton-Jacobi-Bellman (HJB) equation with continuous-time backpropagation and automatic differentiation to compute exact gradients of the value functional from sampled trajectories. Instead of solving the full HJB equation, APIC samples uncontrolled trajectories forward and propagates adjoint co-states backward to compute control gradients efficiently. The controlled dynamics define a Hamiltonian bridge: a physically constrained interpolation between initial and target patterns that respects conservation laws and energetic structure. The framework is demonstrated across canonical systems—navigation on complex landscapes, Allen-Cahn (non-conserved) and Cahn-Hilliard (conserved) phase separation, thin-film droplet self-assembly, reaction-diffusion patterning, and a cell-fate/migration model—showing scalable control and interpretable transport geometries. Conceptually, this links generative modeling and optimal control with variational physics, offering a route to programmable pattern formation in soft matter, active media, and developmental systems and suggesting a path toward inference-in-the-loop and experimental actuation.

Effective wound healing relies on thousands of cells collectively migrating to close the gap. It is increasingly clear that part of this migration is driven by endogenous electrochemical fields which point towards the center of skin wounds and guide collective cell migration through “electrotaxis”. Mounting evidence suggests exogenously applied electric fields can accelerate healing, but progress in this field has been limited by the use of brute-force, global stimulation strategies that are wound agnostic and reflect neither the collective nature of the healing process nor the dynamic nature of the wound geometry. Here, we develop an experimental system to study how monolayer mouse skin responds to spatiotemporally patterned electric fields. We first show that local electrical stimulation can produce near global cell migration responses arising from cell-cell mechanical adhesion. We apply this strategy to 2D circular wounds using a local ring electric field near the wound edge and reveal how to tune the timing of local stimulation to avoid cellular jamming. Finally, we integrate our findings with a biophysics-informed optimal control strategy to tune both when and where the electric field should be applied in time, resulting in dramatic improvements to healing.

We explore the dynamical response of the free surface of an ultra-soft solid driven by a localized moving pressure disturbance. Experiments reveal a steady V-shaped wake analogous to a surface Mach wedge. A simple geometric argument provides a qualitative explanation consistent with observations. A theoretical framework combining elastodynamic, capillary, and gravitational effects yields a generalized dispersion relation that smoothly interpolates between Kelvin's theory of liquid interface wakes and Rayleigh's theory of elastic surface waves. Together, our experiments and theory reveal the existence of a soft wake regime that bridges fluid and solid surface wave physics, offering new routes for probing the dynamics of soft surfaces.

Ducks display a unique and dramatic sexual dimorphism in their vocal organ, the syrinx. Males have a left-sided bulla that is not present in females and that has been long hypothesized to play a role in courtship vocalizations, though this connection has never been tested. The large, hollow morphology of the bulla and its proximity to the sound-producing vocal folds introduce the possibility that it may work as a Helmholtz resonator, which makes it possible to predict the resonance frequencies enhanced by this structure. We found that during early ontogeny, the distribution of energy across the harmonic spectrum of contact calls is not different between males and females. We then used microcomputed tomography (µCT) scans of duck syringes to estimate resonance frequencies of the bullae and compared these with spectral features of their vocalizations. This comparison overall supports the idea that the bulla resonance may specifically enhance aspects of courtship vocalizations, especially in species that have a tonal courtship whistle. This was further supported when we tested the frequencies produced when air was blown through 3D printed bullae. We also saw potential influence of the bulla in non-courtship vocalizations, which could be explored further with a greater understanding of the input of other vocal tract features that influence vocalization. We observed that, in general, excepting the common eider, bulla size shows a weak positive correlation with male bird body mass. This study provides support for the long-held hypothesis that the adult male duck bulla influences resonance frequencies, in particular in courtship vocalizations.

Brains likely evolved from diffuse nerve nets in the Precambrian, but we do not know what the first brains looked like or how they were organized. Acoel worms, the sister lineage to all other animals with brains, offer a window into this transition. We studied the three-banded panther worm Hofstenia miamia, whose brain is diffuse and unlike any previously described: it shows little anatomical or functional regionalization or stereotypy. Worms forage successfully even after large portions of the brain are removed, suggesting most regions can perform most computations. Neural cell type markers are also distributed across the brain with little regionalization. High-resolution studies of hunting reveal that more brain tissue improves performance, but no specific brain region is required. These results lead us to propose that H. miamia’s brain is built from computationally pluripotent ‘tiles’, whose interactions generate coherent behavior. This architecture suggests that early brains arose by condensation of diffuse nerve nets into unregionalized brains, with regionalization evolving secondarily.