mlsanityData to 14 Oct 2026

Learning Object Properties Using Robot Proprioception via Differentiable Robot-Object Interaction

arXiv 2410.03920 ↗First posted 4 Oct 2024cs.RO cs.AI cs.CEcited by 5

Not on the map. Why?

Abstract (arXiv)

Differentiable simulation has become a powerful tool for system identification. While prior work has focused on identifying robot properties using robot-specific data or object properties using object-specific data, our approach calibrates object properties by using information from the robot, without relying on data from the object itself. Specifically, we utilize robot joint encoder information, which is commonly available in standard robotic systems. Our key observation is that by analyzing the robot's reactions to manipulated objects, we can infer properties of those objects, such as inertia and softness. Leveraging this insight, we develop differentiable simulations of robot-object interactions to inversely identify the properties of the manipulated objects. Our approach relies solely on proprioception -- the robot's internal sensing capabilities -- and does not require external measurement tools or vision-based tracking systems. This general method is applicable to any articulated robot and requires only joint position information. We demonstrate the effectiveness of our method on a low-cost robotic platform, achieving accurate mass and elastic modulus estimations of manipulated objects with just a few seconds of computation on a laptop.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Builds on 29 of its 67 references that are in the map. First posted on arXiv 4 Oct 2024.

Builds on

29 of 67 references.

References · 29

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2016SGDR: Stochastic Gradient Descent with Warm Restartscited by 5,045
  3. 2019Learning agile and dynamic motor skills for legged robotscited by 504
  4. 2018Learning Particle Dynamics for Manipulating Rigid Bodies, Deformable Objects, and Fluidscited by 220
  5. 2023PAC-NeRF: Physics Augmented Continuum Neural Radiance Fields for Geometry-Agnostic System Identificationcited by 154
  6. 2019TossingBot: Learning to Throw Arbitrary Objects with Residual Physicscited by 137
  7. 20213D Neural Scene Representations for Visuomotor Controlcited by 111
  8. 2016A Differentiable Physics Engine for Deep Learning in Roboticscited by 108
  9. 2021PlasticineLab: A Soft-Body Manipulation Benchmark with Differentiable Physicscited by 103
  10. 2018ChainQueen: A Real-Time Differentiable Physical Simulator for Soft Roboticscited by 98
  11. 2020NeuralSim: Augmenting Differentiable Simulators with Neural Networkscited by 82
  12. 2021Robot Learning from Randomized Simulations: A Reviewcited by 63
17 more
  1. 2021DiffPD: Differentiable Projective Dynamicscited by 62
  2. 2023Learning Neural Constitutive Laws From Motion Observations for Generalizable PDE Dynamicscited by 57
  3. 2019Sim-to-Real Transfer for Biped Locomotioncited by 56
  4. 2020Visual Grounding of Learned Physical Modelscited by 55
  5. 2021Fast and Feature-Complete Differentiable Physics for Articulated Rigid Bodies with Contactcited by 52
  6. 2021Coupling Vision and Proprioception for Navigation of Legged Robotscited by 51
  7. 2022Differentiable Simulation of Soft Multi-body Systemscited by 39
  8. 2022Iterative Residual Policy: for Goal-Conditioned Dynamic Manipulation of Deformable Objectscited by 37
  9. 2022RISP: Rendering-Invariant State Predictor with Differentiable Simulation and Rendering for Cross-Domain Parameter Estimationcited by 29
  10. 2020Differentiable Physics Models for Real-world Offline Model-based Reinforcement Learningcited by 27
  11. 2020Encoding Physical Constraints in Differentiable Newton-Euler Algorithmcited by 23
  12. 2019Interactive Differentiable Simulationcited by 17
  13. 2021Probabilistic Inference of Simulation Parameters via Parallel Differentiable Simulationcited by 17
  14. 2023Model-Based Control with Sparse Neural Dynamicscited by 16
  15. 2022Inferring Articulated Rigid Body Dynamics from RGBD Videocited by 10
  16. 2023Residual Physics Learning and System Identification for Sim-to-real Transfer of Policies on Buoyancy Assisted Legged Robotscited by 8
  17. 2021A Differentiable Newton-Euler Algorithm for Real-World Roboticscited by 4

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 5.

5 papers.

  1. 2026D-REX: Differentiable Real-to-Sim-to-Real Engine for Learning Dexterous Graspingcited by 3
  2. 2026HALO:Closing Sim-to-Real Gap for Heavy-loaded Humanoid Agile Motion Skills via Differentiable Simulationcited by 1
  3. 2026NeuralActuator: Neural Actuation Modeling for Robot Dynamics and External Force Perceptioncited by 1
  4. 2026FACT: Fidelity-Aware Construction of Articulated Twinscited by 0
  5. 2026DexNinja: Learning Robust Dexterous Cutting Policy with a Real-to-Sim-to-Real Data Enginecited by 0