# Sim-to-Real of Soft Robots with Learned Residual Physics

arXiv: https://arxiv.org/abs/2402.01086 · HTML: https://mlsanity.com/paper/2402.01086/

## Key facts

- Cited by 15 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 2 Feb 2024.

## Built on by (most cited first, 15 of 15)

- [The Reality Gap in Robotics: Challenges, Solutions, and Best Practices](https://mlsanity.com/paper/2510.20808/) (2025)
- [QuasiSim: Parameterized Quasi-Physical Simulators for Dexterous Manipulations Transfer](https://mlsanity.com/paper/2404.07988/) (2024)
- [Accelerated co-design of robots through morphological pretraining](https://mlsanity.com/paper/2502.10862/) (2025)
- [Learning Low-Dimensional Strain Models of Soft Robots by Looking at the Evolution of Their Shape with Application to Model-Based Control](https://mlsanity.com/paper/2411.00138/) (2024)
- [Generalizable and Fast Surrogates: Model Predictive Control of Articulated Soft Robots using Physics-Informed Neural Networks](https://mlsanity.com/paper/2502.01916/) (2025)
- [Learning Soft Robotic Dynamics with Active Exploration](https://mlsanity.com/paper/2510.27428/) (2025)
- [Learning Physics-Guided Residual Dynamics for Deformable Object Simulation](https://mlsanity.com/paper/2607.13451/) (2026)
- [Learning Equivariant Neural-Augmented Object Dynamics From Few Interactions](https://mlsanity.com/paper/2605.02699/) (2026)
- [PhysCoRe: Physics-Corrected Residual World Models for Material-Aware Deformable Dynamics](https://mlsanity.com/paper/2607.20653/) (2026)
- [HDVIO2.0: Wind and Disturbance Estimation with Hybrid Dynamics VIO](https://mlsanity.com/paper/2504.00969/) (2025)
- [Modelling of Underwater Vehicles using Physics-Informed Neural Networks with Control](https://mlsanity.com/paper/2504.20019/) (2025)
- [Bridging the Sim-to-Real Gap with multipanda_ros2: A Real-Time ROS2 Framework for Multimanual Systems](https://mlsanity.com/paper/2602.02269/) (2026)
- [Sharpness-Aware Hybrid Model Learning for Architecture-Agnostic Parameter Estimation](https://mlsanity.com/paper/2602.06837/) (2026)
- [RAFL: Generalizable Sim-to-Real of Soft Robots with Residual Acceleration Field Learning](https://mlsanity.com/paper/2603.22039/) (2026)
- [SoRoMoX: Fast, Differentiable, and Parallelizable Soft Robot Models](https://mlsanity.com/paper/2608.06650/) (2026)

## Abstract (arXiv)

Accurately modeling soft robots in simulation is computationally expensive and commonly falls short of representing the real world. This well-known discrepancy, known as the sim-to-real gap, can have several causes, such as coarsely approximated geometry and material models, manufacturing defects, viscoelasticity and plasticity, and hysteresis effects. Residual physics networks learn from real-world data to augment a discrepant model and bring it closer to reality. Here, we present a residual physics method for modeling soft robots with large degrees of freedom. We train neural networks to learn a residual term -- the modeling error between simulated and physical systems. Concretely, the residual term is a force applied on the whole simulated mesh, while real position data is collected with only sparse motion markers. The physical prior of the analytical simulation provides a starting point for the residual network, and the combined model is more informed than if physics were learned tabula rasa. We demonstrate our method on 1) a silicone elastomeric beam and 2) a soft pneumatic arm with hard-to-model, anisotropic fiber reinforcements. Our method outperforms traditional system identification up to 60%. We show that residual physics need not be limited to low degrees of freedom but can effectively bridge the sim-to-real gap for high dimensional systems.