+
Skip to main content
Springer Nature Link
Log in

ASBI: Leveraging informative real-world data for active black-box simulator tuning

  • Published:
Applied Intelligence Aims and scope Submit manuscript

Abstract

Black-box simulators are widely used in robotics, but optimizing their parameters remains challenging due to inaccessible likelihoods. Simulation-Based Inference (SBI) tackles this issue using simulation-driven approaches, estimating the posterior from offline real observations and forward simulations. However, in black-box scenarios, preparing observations that contain sufficient information for parameter estimation is difficult due to the unknown relationship between parameters and observations. In this work, we present Active Simulation-Based Inference (ASBI), a parameter estimation framework that uses robots to actively collect real-world online data to achieve accurate black-box simulator tuning. Our framework optimizes robot actions to collect informative observations by maximizing information gain, which is defined as the expected reduction in Shannon entropy between the posterior and the prior. While calculating information gain requires the likelihood, which is inaccessible in black-box simulators, our method solves this problem by leveraging Neural Posterior Estimation (NPE), which leverages a neural network to learn the posterior estimator. Three simulation experiments quantitatively verify that our method achieves accurate parameter estimation, with posteriors sharply concentrated around the true parameters. Moreover, we show a practical application using a real robot to estimate the simulation parameters of cubic particles corresponding to two real objects, beads and gravel, with a bucket pouring action.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Subscribe and save

Springer+
from $39.99 /Month
  • Starting from 10 chapters or articles per month
  • Access and download chapters and articles from more than 300k books and 2,500 journals
  • Cancel anytime
View plans

Buy Now

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Algorithm 1
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14
Fig. 15
Fig. 16

Similar content being viewed by others

Explore related subjects

Discover the latest articles, books and news in related subjects, suggested using machine learning.

Data Availability

No datasets were generated or analyzed during the current study.

References

  1. Cranmer K, Brehmer J, Louppe G (2020) The frontier of simulation-based inference. Proc Natl Acad Sci U S A 117(48):30055–30062. https://doi.org/10.1073/pnas.1912789117

    Article  MathSciNet  Google Scholar 

  2. Pina-Otey S, Sánchez F, Gaitan V et al (2020) Likelihood-free inference of experimental neutrino oscillations using neural spline flows. Phys Rev D 101:113001. https://doi.org/10.1103/PhysRevD.101.113001

    Article  Google Scholar 

  3. Lueckmann JM, Gonçalves PJ, Bassetto G et al (2017) Flexible statistical inference for mechanistic models of neural dynamics. In: Advances in neural information processing systems, pp 1289–1299

  4. Vasist M, Rozet F, Absil O et al (2023) Neural posterior estimation for exoplanetary atmospheric retrieval. Astron Astrophys 672:A147. https://doi.org/10.1051/0004-6361/202245263

    Article  Google Scholar 

  5. Dax M, Green SR, Gair J et al (2021) Real-time gravitational wave science with neural posterior estimation. Phys Rev Lett 127:241103. https://doi.org/10.1103/PhysRevLett.127.241103

    Article  Google Scholar 

  6. Khullar G, Nord B, Ćiprijanović A et al (2022) DIGS: deep inference of galaxy spectra with neural posterior estimation. Mach Learn: Sci Technol 3(4):04LT04. https://doi.org/10.1088/2632-2153/ac98f4

    Article  Google Scholar 

  7. Lips T, De Gusseme VL, Wyffels F (2024) Learning keypoints for robotic cloth manipulation using synthetic data. IEEE Robot Autom Lett 9(7):6528–6535. https://doi.org/10.1109/LRA.2024.3405335

    Article  Google Scholar 

  8. Sundaresan P, Grannen J, Thananjeyan B et al (2020) Learning rope manipulation policies using dense object descriptors trained on synthetic depth data. In: 2020 IEEE International Conference on Robotics and Automation (ICRA), pp 9411–9418. https://doi.org/10.1109/ICRA40945.2020.9197121

  9. Mou F, Wang B, Wu D (2022) Learning-based cable coupling effect modeling for robotic manipulation of heavy industrial cables. Sci Rep 12(1):6036. https://doi.org/10.1038/s41598-022-09643-6

    Article  Google Scholar 

  10. Kadokawa Y, Hamaya M, Tanaka K (2023) Learning robotic powder weighing from simulation for laboratory automation. In: 2023 IEEE/RSJ International conference on Intelligent Robots and Systems (IROS), pp 2932–2939. https://doi.org/10.1109/IROS55552.2023.10342463

  11. Egli P, Gaschen D, Kerscher S et al (2022) Soil-adaptive excavation using reinforcement learning. IEEE Robot Autom Lett 7(4):9778–9785. https://doi.org/10.1109/LRA.2022.3189834

    Article  Google Scholar 

  12. Mittal M, Yu C, Yu Q et al (2023) Orbit: a unified simulation framework for interactive robot learning environments. IEEE Robot Autom Lett 8(6):3740–3747. https://doi.org/10.1109/LRA.2023.3270034

    Article  Google Scholar 

  13. Algoryx Simulations (2025) AGX dynamics. https://www.algoryx.se/agx-dynamics. Accessed 06 Mar 2025

  14. Lindley DV (1956) On a measure of the information provided by an experiment. Ann Math Stat 27(4):986–1005. https://doi.org/10.1214/aoms/1177728069

    Article  MathSciNet  MATH  Google Scholar 

  15. Papamakarios G, Murray I (2016) Fast \(\epsilon\)-free inference of simulation models with Bayesian conditional density estimation. In: Advances in neural information processing systems, pp 1028–1036

  16. Greenberg D, Nonnenmacher M, Macke J (2019) Automatic posterior transformation for likelihood-free inference. In: Proceedings of the 36th international conference on machine learning, pp 2404–2414

  17. Kleinegesse S, Drovandi C, Gutmann MU (2021) Sequential Bayesian experimental design for implicit models via mutual information. Bayesian Anal 16(3):773–802. https://doi.org/10.1214/20-BA1225

    Article  MathSciNet  MATH  Google Scholar 

  18. Pritchard JK, Seielstad MT, Perez-Lezaun A et al (1999) Population growth of human y chromosomes: a study of y chromosome microsatellites. Mol Biol Evol 16(12):1791–1798. https://doi.org/10.1093/oxfordjournals.molbev.a026091

    Article  Google Scholar 

  19. Beaumont MA, Zhang W, Balding DJ (2002) Approximate Bayesian computation in population genetics. Genetics 162(4):2025–2035. https://doi.org/10.1093/genetics/162.4.2025

    Article  Google Scholar 

  20. Marjoram P, Molitor J, Plagnol V et al (2003) Markov chain Monte Carlo without likelihoods. Proc Natl Acad Sci U S A 100(26):15324–15328. https://doi.org/10.1073/pnas.0306899100

    Article  Google Scholar 

  21. Bonassi FV, West M (2015) Sequential Monte Carlo with adaptive weights for approximate Bayesian computation. Bayesian Anal 10(1):171–187. https://doi.org/10.1214/14-BA891

    Article  MathSciNet  MATH  Google Scholar 

  22. Possas R, Barcelos L, Oliveira R et al (2020) Online BayesSim for combined simulator parameter inference and policy improvement. In: 2020 IEEE/RSJ International conference on Intelligent Robots and Systems (IROS), pp 5445–5452. https://doi.org/10.1109/IROS45743.2020.9341401

  23. Muratore F, Gruner T, Wiese F et al (2022) Neural posterior domain randomization. In: Proceedings of the 5th conference on robot learning, pp 1532–1542

  24. Myung JI, Cavagnaro DR, Pitt MA (2013) A tutorial on adaptive design optimization. J Math Psychol 57(3):53–67. https://doi.org/10.1016/j.jmp.2013.05.005

    Article  MathSciNet  MATH  Google Scholar 

  25. Dushenko S, Ambal K, McMichael RD (2020) Sequential Bayesian experiment design for optically detected magnetic resonance of nitrogen-vacancy centers. Phys Rev Appl 14:054036. https://doi.org/10.1103/PhysRevApplied.14.054036

    Article  Google Scholar 

  26. Rainforth T, Foster A, Ivanova DR et al (2024) Modern Bayesian experimental design. Stat Sci 39(1):100–114. https://doi.org/10.1214/23-STS915

    Article  MathSciNet  MATH  Google Scholar 

  27. Ryan EG, Drovandi CC, McGree JM et al (2016) A review of modern computational algorithms for Bayesian optimal design. Int Stat Rev 84(1):128–154. https://doi.org/10.1111/insr.12107

    Article  MathSciNet  MATH  Google Scholar 

  28. Saal HP, Ting JA, Vijayakumar S (2010) Active estimation of object dynamics parameters with tactile sensors. In: 2010 IEEE/RSJ International conference on Intelligent Robots and Systems (IROS), pp 916–921. https://doi.org/10.1109/IROS.2010.5649191

  29. Cooper M, McGree J, Molloy TL et al (2021) Bayesian experimental design with application to dynamical vehicle models. IEEE Trans Robot 37(5):1844–1851. https://doi.org/10.1109/TRO.2021.3063977

    Article  Google Scholar 

  30. Dutta A, Burdet E, Kaboli M (2023) Push to know! - visuo-tactile based active object parameter inference with dual differentiable filtering. In: 2023 IEEE/RSJ International conference on Intelligent Robots and Systems (IROS), pp 3137–3144. https://doi.org/10.1109/IROS55552.2023.10341832

  31. Margolis GB, Fu X, Ji Y et al (2023) Learning to see physical properties with active sensing motor policies. In: Proceedings of the 7th conference on robot learning, pp 2537–2548

  32. Memmel M, Wagenmaker A, Zhu C et al (2024) ASID: active exploration for system identification in robotic manipulation. arXiv:2404.12308. https://arxiv.org/abs/2404.12308

  33. Thomas O, Dutta R, Corander J et al (2022) Likelihood-free inference by ratio estimation. Bayesian Anal 17(1):1–31. https://doi.org/10.1214/20-BA1238

    Article  MathSciNet  Google Scholar 

  34. Ramos F, Possas R, Fox D (2019) BayesSim: Adaptive domain randomization via probabilistic inference for robotics simulators. In: Proceedings of robotics: science and systems XV. https://doi.org/10.15607/rss.2019.xv.029

  35. Papamakarios G, Sterratt D, Murray I (2019) Sequential neural likelihood: fast likelihood-free inference with autoregressive flows. In: Proceedings of the twenty-second international conference on artificial intelligence and statistics, pp 837–848

  36. Ward D, Cannon P, Beaumont M et al (2022) Robust neural posterior estimation and statistical model criticism. In: Advances in neural information processing systems, pp 33845–33859

  37. Tian WL, Gao P, Liu X et al (2025) Toward adaptive meta-gradient adversarial examples for visual tracking. IEEE Trans Reliab 1–14. https://doi.org/10.1109/TR.2025.3569828

  38. Xu L, Gao P, Tang WJ et al (2025) Towards effective and efficient adversarial defense with diffusion models for robust visual tracking. Inf Fusion 124:103384. https://doi.org/10.1016/j.inffus.2025.103384

    Article  Google Scholar 

  39. Shahriari B, Swersky K, Wang Z et al (2016) Taking the human out of the loop: a review of Bayesian optimization. Proc IEEE 104(1):148–175. https://doi.org/10.1109/JPROC.2015.2494218

    Article  Google Scholar 

  40. Nguyen V (2019) Bayesian optimization for accelerating hyper-parameter tuning. In: 2019 IEEE Second international conference on Artificial Intelligence and Knowledge Engineering (AIKE), pp 302–305, https://doi.org/10.1109/AIKE.2019.00060

  41. Rumelhart DE, Hinton GE, Williams RJ (1986) Learning representations by back-propagating errors. Nature 323(6088):533–536. https://doi.org/10.1038/323533a0

    Article  Google Scholar 

  42. Hecht-Nielsen (1989) Theory of the backpropagation neural network. In: Proceedings of the international 1989 joint conference on neural networks, pp 593–605. https://doi.org/10.1109/IJCNN.1989.118638

  43. Tao H, Zheng Y, Wang Y et al (2024) Enhanced feature extraction yolo industrial small object detection algorithm based on receptive-field attention and multi-scale features. Meas Sci Technol 35(10):105023. https://doi.org/10.1088/1361-6501/ad633d

    Article  Google Scholar 

  44. Tao H, Huang Z, Wang Y et al (2025) Efficient feature fusion network for small objects detection of traffic signs based on cross-dimensional and dual-domain information. Meas Sci Technol 36(3):035004. https://doi.org/10.1088/1361-6501/adb2ad

    Article  Google Scholar 

  45. Sun Y, Tao H, Stojanovic V (2025) End-to-end multi-scale residual network with parallel attention mechanism for fault diagnosis under noise and small samples. ISA Trans 157:419–433. https://doi.org/10.1016/j.isatra.2024.12.023

    Article  Google Scholar 

  46. Li Q, Kroemer O, Su Z et al (2020) A review of tactile information: perception and action through touch. IEEE Trans Rob 36(6):1619–1634. https://doi.org/10.1109/TRO.2020.3003230

    Article  Google Scholar 

  47. Kaboli M, Yao K, Cheng G (2016) Tactile-based manipulation of deformable objects with dynamic center of mass. In: 2016 IEEE-RAS 16th international conference on humanoid robots (Humanoids), pp 752–757. https://doi.org/10.1109/HUMANOIDS.2016.7803358

  48. Puentes K, Morales L, Pozo-Espin DF et al (2024) Enhancing control systems with neural network-based intelligent controllers. Emerg Sci J 8(4):1243–1261. https://doi.org/10.28991/esj-2024-08-04-01

    Article  Google Scholar 

  49. Huajun Z, Xinchi T, Hang G et al (2020) The parameter identification of the autonomous underwater vehicle based on multi-innovation least squares identification algorithm. Int J Adv Robot Syst 17(2):1729881420921016. https://doi.org/10.1177/1729881420921016

    Article  Google Scholar 

  50. Rasul T, Mukherjee K (2024) Data-driven approach for parameter estimation and control of an autonomous underwater vehicle. JEMS Maritime Sci 12(2):144–155. https://doi.org/10.4274/jems.2024.10438

    Article  Google Scholar 

  51. Vargas M, Vasquez Y, Barra D et al (2024) Elbow-hand robotic exoskeletons for active and passive rehabilitation on post-stroke patients: a bioengineering review. HighTech Innov J 5(4):1170–1190. https://doi.org/10.28991/hij-2024-05-04-020

    Article  Google Scholar 

  52. Cornejo J, Cornejo J, Vargas M et al (2024) SY-MIS project: biomedical design of endo-robotic and laparoscopic training system for surgery on the earth and space. Emerg Sci J 8(2):372–393. https://doi.org/10.28991/esj-2024-08-02-01

    Article  Google Scholar 

Download references

Acknowledgements

This work is supported by JST [Moonshot Research and Development], Grant Number [JPMJMS2032].

Author information

Authors and Affiliations

  1. Graduate School of Science and Technology, Nara Institute of Science and Technology, Nara, Japan

    Gahee Kim & Takamitsu Matsubara

Authors
  1. Gahee Kim
    View author publications

    Search author on:PubMed Google Scholar

  2. Takamitsu Matsubara
    View author publications

    Search author on:PubMed Google Scholar

Contributions

Gahee Kim: Conceptualization, Methodology, Software, Experiment, Analysis, Visualization, Writing. Takamitsu Matsubara: Conceptualization, Writing - Review & Editing, Supervision, Project administration.

Corresponding author

Correspondence to Gahee Kim.

Ethics declarations

Competing Interests

The authors have no competing interests to declare.

Ethical and Informed Consent for Data Used

This article does not contain any studies with human participants or animals.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Kim, G., Matsubara, T. ASBI: Leveraging informative real-world data for active black-box simulator tuning. Appl Intell 55, 1028 (2025). https://doi.org/10.1007/s10489-025-06934-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • Version of record:

  • DOI: https://doi.org/10.1007/s10489-025-06934-z

Keywords

点击 这是indexloc提供的php浏览器服务，不要输入任何密码和下载