Soft yet Effective Robots via Holistic Co-Design

TL;DR

The paper tackles the challenge of designing safe, reliable soft robots by moving beyond traditional sequential design to a holistic co-design framework that jointly optimizes body and brain while accounting for manufacturability, safety, and regulatory constraints. It proposes a multi-value, computation-friendly approach that leverages reduced-order models, surrogate metrics, probabilistic evaluation, and targeted prototyping to enable global design-space exploration and mitigate sim-to-real gaps. Key contributions include a formal definition of computational co-design for soft robots, a control-oriented reduced-order modeling strategy, surrogate and probabilistic metrics, and mechanisms for reproducibility and stakeholder involvement. The framework aims to improve robustness, safety, and public trust in soft robotics, enabling more widespread adoption in human-centered applications.

Abstract

Soft robots promise inherent safety via their material compliance for seamless interactions with humans or delicate environments. Yet, their development is challenging because it requires integrating materials, geometry, actuation, and autonomy into complex mechatronic systems. Despite progress, the field struggles to balance task-specific performance with broader factors like durability and manufacturability - a difficulty that we find is compounded by traditional sequential design processes with their lack of feedback loops. In this perspective, we review emerging co-design approaches that simultaneously optimize the body and brain, enabling the discovery of unconventional designs highly tailored to the given tasks. We then identify three key shortcomings that limit the broader adoption of such co-design methods within the soft robotics domain. First, many rely on simulation-based evaluations focusing on a single metric, while real-world designs must satisfy diverse criteria. Second, current methods emphasize computational modeling without ensuring feasible realization, risking sim-to-real performance gaps. Third, high computational demands limit the exploration of the complete design space. Finally, we propose a holistic co-design framework that addresses these challenges by incorporating a broader range of design values, integrating real-world prototyping to refine evaluations, and boosting efficiency through surrogate metrics and model-based control strategies. This holistic framework, by simultaneously optimizing functionality, durability, and manufacturability, has the potential to enhance reliability and foster broader acceptance of soft robotics, transforming human-robot interactions.

Figures (4)

• Figure 1: Holistic Co-Design of Soft Robots. The five pillars of holistic co‑design are (1) Incorporating comprehensive design requirements and values directly into multi‑objective optimization; (2) Efficiently exploring the full design space by melding design priors (e.g., biological inspirations, existing solutions) with computationally efficient co‑design routines; (3) Explicitly accounting for design realization (e.g., prototyping and testing) via a probabilistic treatment of evaluation metrics and by formalizing the refinement‑vs‑realization trade‑off—enabling targeted prototyping to reduce metric uncertainty and narrow the sim‑to‑real gap; (4) Fostering cross‑disciplinary collaboration and involving all relevant stakeholders in defining design values and providing iterative feedback; (5) Ensuring auditability to preserve design knowledge and guarantee reproducibility. These pillars are inherently interconnected (see green arrows)—for example, stakeholders and the design team co‑define design requirements, certain design values (e.g., HRI) demand evaluation in the real world (i.e., enabled by realization), and both the exploration of the design space and the fulfillment of design requirements remain fully auditable.
• Figure 2: The traditional soft robot design process vs. holistic co-design.Left: A traditional sequential design cycle van2020delft applied to soft robots. This sequential workflow lacks iterative feedback loops, leading to information silos and preventing regular, bidirectional sharing of insights and data across the development team and the relevant stakeholders. Right: A holistic co-design framework that complements computational refinement with prototyping across various levels of design readiness - visualized as "onion" layers in this graphic. Solid yellow lines denote the prototyping pathway, while dashed orange arrows illustrate how empirical findings continuously inform and update the estimates $\hat{c}_j$ for the real‑world design values $c_j$.
• Figure 3: A Framework for Efficient Computational Co-Design of Soft Robots. We sample reduced-order design parameters $x$ from an initial distribution that can include design priors from biological inspiration mazzolai2020visionchen2020designlaschi2024bioinspiration, known mechanisms, or existing soft robot designs. This design space—either learned or explicitly defined by physical or geometric values (i.e., size optimization)—is translated by a design decoder into a detailed robot body description (such as a 3D mesh with sensor and actuator placements) that provides immediate manufacturability feedback. A reduced-order model is then derived alora2023datastolzle2024inputvaladas2025learningalkayas2025soft or learned to efficiently assess workspace, open-loop compliance, and embodied intelligence cianchetti2021embodiedmengaldo2022concisevihmar2023measure, offering inexpensive feedback to the optimizer. Similarly, perception and control systems can be developed using this model (via model-based control della2023model, MPC alora2023data, or differentiable physics spielberg2019learningwang2023softzoo), while observability and controllability are evaluated without costly simulations. Finally, closed-loop simulations (low-fidelity using the reduced-order model or high-fidelity FEM-based coevoet2017software) assess integrated performance metrics, which in turn optimize the sampling distribution, decoder, and all relevant system parameters. References in graphic: (1): navez2024contributions, (2): bhatia2021evolution, (3): wang2023softzoo, (4): wang2024diffusebot, (5): song2024morphvae, (6): sutton1998reinforcement, (7): garnett2023bayesian, (8): medvet2022impact, (9): guan2023trimmed, (10): armanini2023soft, (11): valadas2025learning, (12): alkayas2025soft, (13): menager2023direct, (14): alora2023data, (15): spielberg2021co, (16): junge2022leveraging, (17): zheng2019controllability, (18): amehri2022workspace, (19): cianchetti2021embodied, (20): mengaldo2022concise, (21): vihmar2023measure, (22) spielberg2019learning, (23): della2023model.
• Figure 4: Refinement vs. Realization.Top Left: This panel shows how refinement affects the design sampling distribution $\vartheta(x)$. Unlike (most) conventional co‑design methods that sparsely explore the design space, we condition $\vartheta(x)$ on design priors (e.g., biological inspiration and existing soft robot designs) and iteratively update it via the co‑design optimizer until the optimum is reached. Bottom Left: Here, we demonstrate how iterative realization refines probabilistic evaluation metrics. Rather than using deterministic metrics that ignore inherent simulation uncertainty, we define a probabilistic distribution $\hat{m}_j(\hat{c}_j, x)$ for each metric, with purposeful prototyping reducing the uncertainty over iterations. Right: This panel examines the tradeoff between refinement and realization. We define the realization ratio $\mathrm{RealR}$ as the fraction of realization iterations relative to the total iterations and the sim‑to‑real error as $\lVert c_j - \mathbb{E}[\hat{c}_j] \rVert$. We then plot the evolution of real-world cost $c_j$ versus sim‑to‑real error as iterations increase for a fixed $\mathrm{RealR}$: when $\lim{\mathrm{RealR} \to 0}$, designs are optimized computationally with realization occurring only at the end, while $\mathrm{RealR}=1$ means only the confidence in the evaluation metrics is increased. The optimal ratio $\mathrm{RealR}^*$ efficiently minimizes the real-world cost to its optimum $\min_x c_j(x)$ via refinement and reduces the sim‑to‑real error to zero by updating $\hat{m}_j(\hat{c}_j, x)$ via realization.