A Category-Theoretic Framework from Biological Mechanics to Engineered Stimulus-Response Systems

Abstract: Natural materials achieve adaptive behavior through hierarchical organization and coupled mechanisms across scales. Their translation into engineering, however, remains largely heuristic. What is missing is a formal translation framework that carries biological design logic into engineered realization while preserving physical consistency across levels of abstraction. Here we present a category theoretic compositional framework for verified nature-derived design. The framework defines a category of stimulus response dynamical systems with natural and artificial subcategories. It introduces a structure preserving implementation functor from biological mechanics to engineered systems. It also formalizes a machine agnostic specification layer that links behavioral intent to executable fabrication programs. We instantiate the framework on the hygromorphic pinecone hierarchy as a representative biological case. We implement the full pipeline in Grasshopper, where formal specifications are translated into modular parametric scripts that preserve the compositional structure of the model. The resulting designs are fabricated by fused filament fabrication, evaluated experimentally, and tested against model predictions derived from the pipeline. The current implementation generates four actuator classes spanning two stimulus types and two kinematic responses. One actuator arises purely through composition from previously validated components, without additional manual derivation. The results show that compositionality can function not just as a descriptive language, but as a generative and system level verifiable method for mechanical material design. More broadly, the work provides a concrete route for embedding formal multiscale reasoning within increasingly computational, generative, and physics-driven design workflows.

• The paper presents a compositional category-theoretic framework that rigorously maps biological stimulus-response hierarchies to engineered systems.
• It introduces key functors (𝓕, 𝓔) and categories (Dyn, Nat, Art, Spec) that ensure global behavioral fidelity across multi-scale designs.
• Experimental validation on pinecone hygromorphism demonstrates precise actuator performance and consistency from model prediction to fabrication.

A Category-Theoretic and Compositional Framework Bridging Biological Hierarchy to Engineered Stimulus-Response Systems

Introduction and Motivation

This work introduces a mathematically rigorous, category-theoretic framework to translate stimulus-response mechanisms from biological hierarchies to engineering systems. The authors address a fundamental challenge: while biological materials often function via compositional, multiscale mechanisms, conventional biomimetic engineering typically relies on heuristic or analogy-driven strategies that fail to guarantee global behavioral fidelity when complexity grows. The proposed approach formalizes this translation using compositional category theory, enabling the construction of engineered systems that are hierarchically, dynamically, and structurally isomorphic to their biological counterparts—with correctness and behavioral preservation verifiable by construction.

Formal Framework: $Dyn$, $Nat$, $Art$, $Spec$

The technical backbone of the work is the introduction of the category $Dyn$ (stimulus-response dynamical systems), with $Nat$ and $Art$ as full subcategories representing natural (biological) and artificial (engineered) systems, respectively. Objects of $Dyn$ encapsulate a triplet: state space $X$, stimulus/environment space $E$, and a governing evolution law $Nat$0, such that system dynamics are given by $Nat$1. Morphisms (assemblies, reductions) are smooth maps between these objects that satisfy a strong simulation condition—ensuring that coarse-grained state and stimulus evolution are consistent with their fine-grained originals.

The category-theoretic apparatus built atop $Nat$2 includes:

• Implementation functor $Nat$3: A structure-preserving translation mapping a biophysical hierarchy in $Nat$4 to its engineered analog in $Nat$5, retaining compositional logic but substituting material and process parameters.
• Specification space $Nat$6 and projection $Nat$7: $Nat$8 encodes the machine-agnostic fabrication specification—capturing the many-to-one mapping from process parameters to functional targets. The projection $Nat$9 defines the set of all process schedules that yield the same engineered behavioral target.
• Compilation functor $Art$0: Translates verified, machine-agnostic specifications into executable machine code (e.g., G-code for FFF).

This architecture not only enforces strict compositionality across assembly but enables systematic, verifiable design-space exploration.

Biological Case Study: The Pinecone Hygromorphic Hierarchy

The authors instantiate the framework on the pinecone's humidity-driven actuation mechanism, a canonical multiscale biological system for stimulus-response—modeling it across a nested fiber, lamina, tissue, element, and organ hierarchy in $Art$1.

Figure 1: The pinecone hierarchy in Nat, with dynamic actuation, hierarchical organization across scales, and a formalization as a chain of objects linked by assembly and reduction morphisms.

Key technical contributions for each scale include:

• Fiber scale: Moisture-driven anisotropic swelling modeled via relaxation dynamics for fiber bundles.
• Lamina scale: Assembly morphism averaging fiber states; inherited principal strain response.
• Tissue scale: Multi-lamina composites with differential strain (governed by orientation angles) serving as the source for through-thickness mismatches.
• Element scale: Application of classical beam theory (Timoshenko) to translate strain mismatch into curvature.
• Organ scale: Aggregation of elements into macroscale deformation and actuation (e.g., pinecone opening angle).

Reductions (observable extractions) and assembly morphisms (compositions) ensure the simulation condition and compositional closure at every interface.

Figure 3: Fiber-to-lamina assembly via averaging morphism $Art$2 over $Art$3 fibers, extracting effective lamina observables.

Figure 2: Lamina-to-tissue assembly by stacking $Art$4 laminae with fiber orientations, with reduction extracting macroscopic strain mismatch.

Engineering Translation and Isomorphic Construction

The implementation functor $Art$5 is realized by defining an isomorphic engineered hierarchy within $Art$6—here based on 4D-printed, anisotropic bilayer composites fabricated via fused filament fabrication (FFF).

Figure 4: Top: Biological (Nat) hierarchy; Bottom: Engineered (Art) hierarchy. Vertical arrows depict the functorial translation $Art$7 mapping objects, assemblies, and reductions between the two.

This translation preserves:

• State/Stimulus Structure: State and input spaces are matched in type and organization.
• Dynamical Evolution: Evolution laws have identical form, differing only in material and process parameterization.
• Compositionality: All assembly and reduction morphisms in Nat have direct analogs in Art, ensuring simulation conditions are satisfied at each scale.

This enables direct propagation of verified design logic from biology to fabrication—irrespective of the physical substrate.

Specification and Execution: From Behavioral Target to Fabrication

The framework extends compositionality to manufacturing via the category $Art$8, capturing the multi-faceted mapping from design intent to executable machine programs. The projection $Art$9 ensures that all program variations within an admissible process window (e.g., infill, print speed, nozzle temperature) yield indistinguishable behavioral targets in Art, providing a rigorous means to traverse the fabrication design-space while guaranteeing physical consistency.

Figure 7: Workflow implementation in Grasshopper. Biological and engineered material properties, stimulus collectors, and hierarchical governing equations are linked through a modular, categorical computational graph; output path runs from behavioral target through fabrication specification to machine code.

The entire chain—Nat $Spec$0 Art $Spec$1 Spec $Spec$2 Comp—is operationalized in a parameterized, modular pipeline using Grasshopper (Rhinoceros 3D), ensuring every step retains compositional structure and functional verifiability.

Empirical Validation and Generativity

A central hypothesis—which the paper claims and demonstrates both theoretically and practically—is that compositionality is not merely descriptive but generative. The pipeline is instantiated on four actuator classes, spanning two stimulus types (hygroscopic, thermal) and two kinematic responses (bending, twisting). Only two local variations are introduced: the fiber-scale stimulus law and the tissue-scale reduction (observable extraction). The remaining architecture, functors, and computational pipeline are entirely unchanged.

• Notably, the "thermal twisting" actuator class is **not explicitly constructed**; it arises as an automatic composition of previously validated modules (thermal fiber scale and twisting reduction), highlighting the exponential growth of accessible design space with the number of such modules.

Quantitative agreement is reported between model predictions and experimental measurements for all four actuator classes; every design, fabricated on first attempt with generated G-code, exhibits targeted actuation modes and predicted geometries.

Figure 9: Predicted and observed actuation for all four actuator classes, organized by stimulus type (columns) and response (rows); left: pipeline-computed deformations, right: experimental realization.

Implications and Future Directions

The presented methodology has significant theoretical and practical implications:

• Rigorous Compositional Verification: The framework ensures that behavioral correctness is guaranteed by compositional structure—if all local morphisms are validated, any composite structure is globally valid by construction.
• Generative Design-Space Enlargement: The verification burden scales linearly with the number of independent modules, while the number of constructible devices (via recombination and composition) scales combinatorially.
• Seamless Integration with Generative AI: The structure of $Spec$3, $Spec$4, $Spec$5, and $Spec$6 aligns naturally with graph- and ontology-based reasoning in AI-driven materials discovery [Buehler2024, Buehler2025], enabling rapid, verifiable design generation and search.

Practical extensions are anticipated in several directions, including incorporation of uncertainty propagation [Huang2026], more complex and nonlinear material laws, and multi-agent or neurosymbolic AI systems for automated proposal and verification of new categorical modules.

Conclusion

This work articulates a unified, end-to-end categorical method for translating biological hierarchy and logic into engineered, fabricable, and verifiable stimulus-response systems. The framework's core strength lies in its ability to make the compositional structure explicit and operational—supporting both rigorous verification and generative expansion of the design space. The research lays a formal foundation for integrating rigorous physical modeling, compositional logic, and modern generative AI in materials design, and has clear practical ramifications for the automation and reliability of nature-derived engineering.

---

References

See (2604.26367) for full bibliography.