{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,18]],"date-time":"2026-07-18T05:16:00Z","timestamp":1784351760600,"version":"3.55.0"},"reference-count":61,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,2,23]],"date-time":"2026-02-23T00:00:00Z","timestamp":1771804800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:p>Physics-aware machine learning integrates domain-specific physical knowledge into machine learning models, leading to the development of physics-informed neural networks (PINNs). PINNs embed physical laws directly into the learning process, enabling interpretable and physically consistent solutions to complex problems. However, the practical use of PINNs presents challenges and their applications are complex. Therefore, in this paper, we demonstrate the implementation of PINNs for systems of ordinary differential equations (ODEs), an area that is often overlooked by the physics community, which typically focuses on partial differential equations. We discuss two key challenges: the inverse problem, which involves estimating unknown parameters of ODEs, and the forward problem, which provides an approximate solution to ODEs. To provide practical insights into PINNs, we present two case studies based on a Python implementation using DeepXDE. Drawing on these studies, we discuss key challenges and identify promising directions for future research in PINN-based implementation frameworks.<\/jats:p>","DOI":"10.3389\/frai.2026.1717117","type":"journal-article","created":{"date-parts":[[2026,2,23]],"date-time":"2026-02-23T07:02:21Z","timestamp":1771830141000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":4,"title":["Implementing physics-informed neural networks with deep learning for differential equations"],"prefix":"10.3389","volume":"9","author":[{"given":"Frank","family":"Emmert-Streib","sequence":"first","affiliation":[{"name":"Predictive Society and Data Analytics Lab, Faculty of Information Technology and Communication Sciences, Tampere University","place":["Tampere, Finland"]},{"name":"College of Health and Life Sciences, Hamad Bin Khalifa University","place":["Doha, Qatar"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shailesh","family":"Tripathi","sequence":"additional","affiliation":[{"name":"Josef Ressel Centre for Data-Driven Business Model Innovation, University of Applied Sciences Upper Austria","place":["Steyr, Austria"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Amer","family":"Farea","sequence":"additional","affiliation":[{"name":"Predictive Society and Data Analytics Lab, Faculty of Information Technology and Communication Sciences, Tampere University","place":["Tampere, Finland"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Andreas","family":"Holzinger","sequence":"additional","affiliation":[{"name":"Human-Centered AI Lab, Institute of Forest Engineering, Department of Ecosystem Management, Climate and Biodiversity, BOKU University","place":["Vienna, Austria"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,2,23]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2510.05433","article-title":"Physics-informed machine learning in biomedical science and engineering","author":"Ahmadi","year":"2025","journal-title":"arXiv"},{"key":"B2","doi-asserted-by":"publisher","first-page":"219","DOI":"10.1016\/j.biosystemseng.2023.04.012","article-title":"A novel physics-informed neural networks approach (pinn-mt) to solve mass transfer in plant cells during drying","volume":"230","author":"Batuwatta-Gamage","year":"2023","journal-title":"Biosyst. Eng"},{"key":"B3","doi-asserted-by":"publisher","first-page":"100107","DOI":"10.1016\/j.jcmds.2024.100107","article-title":"Impact of collocation point sampling techniques on pinn performance in groundwater flow predictions","volume":"14","author":"Bauduin","year":"2025","journal-title":"J. Comput. Mathem. Data Sci"},{"key":"B4","doi-asserted-by":"publisher","first-page":"1118","DOI":"10.1002\/bies.10189","article-title":"Modeling transcriptional regulatory networks","volume":"24","author":"Bolouri","year":"2002","journal-title":"BioEssays"},{"key":"B5","doi-asserted-by":"crossref","DOI":"10.1007\/978-1-4939-9828-9","volume-title":"Mathematical Models in Epidemiology","author":"Brauer","year":"2019"},{"key":"B6","doi-asserted-by":"publisher","first-page":"1727","DOI":"10.1007\/s10409-021-01148-1","article-title":"Physics-informed neural networks (pinns) for fluid mechanics: a review","volume":"37","author":"Cai","year":"2021","journal-title":"Acta Mechanica Sinica"},{"key":"B7","first-page":"1","article-title":"\u201cNeural ordinary differential equations,\u201d","author":"Chen","year":"2018","journal-title":"Advances in Neural Information Processing Systems (NeurIPS 2018)"},{"key":"B8","unstructured":"\u201cNeural ordinary differential equations,\u201d\n          \n          \n            \n              Chen\n              R. T. Q.\n            \n            \n              Rubanova\n              Y.\n            \n            \n              Bettencourt\n              J.\n            \n            \n              Duvenaud\n              D. K.\n            \n          \n          Advances in Neural Information Processing Systems\n          \n          2018"},{"key":"B9","doi-asserted-by":"publisher","first-page":"1420","DOI":"10.1109\/TETCI.2023.3298535","article-title":"A generative model for surrogates of spatial-temporal wildfire nowcasting","volume":"7","author":"Cheng","year":"2023","journal-title":"IEEE Trans. Emerg. Topics Comput. Intellig"},{"key":"B10","first-page":"23761","article-title":"\u201cSeparable physics-informed neural networks,\u201d","author":"Cho","year":"2023","journal-title":"Advances in Neural Information Processing Systems"},{"key":"B11","doi-asserted-by":"publisher","first-page":"88","DOI":"10.1007\/s10915-022-01939-z","article-title":"Scientific machine learning through physics\u2013informed neural networks: where we are and what's next","volume":"92","author":"Cuomo","year":"2022","journal-title":"J. Scientif. Comp"},{"key":"B12","unstructured":""},{"key":"B13","unstructured":""},{"key":"B14","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2305.01240","article-title":"Convergence and error analysis of pinns","author":"Doum\u00e8che","year":"2023","journal-title":"arXiv"},{"key":"B15","doi-asserted-by":"publisher","first-page":"1534","DOI":"10.3390\/ai5030074","article-title":"Understanding physics-informed neural networks: techniques, applications, trends, and challenges","volume":"5","author":"Farea","year":"2024","journal-title":"AI"},{"key":"B16","doi-asserted-by":"publisher","first-page":"115771","DOI":"10.1016\/j.cma.2022.115771","article-title":"Physics-data combined machine learning for parametric reduced-order modelling of nonlinear dynamical systems in small-data regimes","volume":"404","author":"Fu","year":"2023","journal-title":"Comp. Methods Appl. Mechan. Eng"},{"key":"B17","doi-asserted-by":"publisher","first-page":"83","DOI":"10.1016\/S0304-3800(98)00136-7","article-title":"Growth models and their use in ecological modelling: an application to a fish population","volume":"113","author":"Gamito","year":"1998","journal-title":"Ecol. Model"},{"key":"B18","doi-asserted-by":"publisher","first-page":"139","DOI":"10.1038\/s41540-024-00460-3","article-title":"Robust parameter estimation and identifiability analysis with hybrid neural ordinary differential equations in computational biology","volume":"10","author":"Giampiccolo","year":"2024","journal-title":"NPJ Syst. Biol. Appl"},{"key":"B19","doi-asserted-by":"publisher","first-page":"62199","DOI":"10.1109\/ACCESS.2024.3395532","article-title":"Human-centered ai in smart farming: towards agriculture 5.0","volume":"12","author":"Holzinger","year":"2024","journal-title":"IEEE Access"},{"key":"B20","doi-asserted-by":"publisher","first-page":"106916","DOI":"10.1016\/j.knosys.2021.106916","article-title":"Classification by ordinal sums of conjunctive and disjunctive functions for explainable ai and interpretable machine learning solutions","volume":"220","author":"Hudec","year":"2021","journal-title":"Knowl. Based Syst"},{"key":"B21","volume-title":"Mathematical Modeling in Systems Biology: an Introduction","author":"Ingalls","year":"2013"},{"key":"B22","doi-asserted-by":"publisher","first-page":"51","DOI":"10.1016\/j.jtbi.2018.02.027","article-title":"Extended logistic growth model for heterogeneous populations","volume":"445","author":"Jin","year":"2018","journal-title":"J. Theoret. Biol"},{"key":"B23","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2511.11734","article-title":"Physics-informed neural odes with scale-aware residuals for learning stiff biophysical dynamics","author":"Kainth","year":"2025","journal-title":"arXiv"},{"key":"B24","doi-asserted-by":"publisher","first-page":"422","DOI":"10.1038\/s42254-021-00314-5","article-title":"Physics-informed machine learning","volume":"3","author":"Karniadakis","year":"2021","journal-title":"Nat. Rev. Phys"},{"key":"B25","doi-asserted-by":"publisher","first-page":"109358","DOI":"10.1016\/j.cpc.2024.109358","article-title":"A python tool for parameter estimation of \u201cblack box\u201d macro-and micro-kinetic models with bayesian optimization\u2013petboa","volume":"306","author":"Kasiraju","year":"2025","journal-title":"Comp. Phys. Commun"},{"key":"B26","doi-asserted-by":"publisher","first-page":"9","DOI":"10.1063\/5.0060697","article-title":"Stiff neural ordinary differential equations","volume":"31","author":"Kim","year":"2021","journal-title":"Chaos: Interdiscip. J. Nonlinear Sci"},{"key":"B27","doi-asserted-by":"publisher","first-page":"1331","DOI":"10.1007\/s12206-021-0342-5","article-title":"Knowledge integration into deep learning in dynamical systems: an overview and taxonomy","volume":"35","author":"Kim","year":"2021","journal-title":"J. Mech. Sci. Technol"},{"key":"B28","doi-asserted-by":"publisher","first-page":"e1008462","DOI":"10.1371\/journal.pcbi.1008462","article-title":"Biologically-informed neural networks guide mechanistic modeling from sparse experimental data","volume":"16","author":"Lagergren","year":"2020","journal-title":"PLoS Comput. Biol"},{"key":"B29","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2404.07662","article-title":"Pinnacle: Pinn adaptive collocation and experimental points selection","author":"Lau","year":"2024","journal-title":"arXiv"},{"key":"B30","doi-asserted-by":"publisher","first-page":"140","DOI":"10.3390\/bdcc6040140","article-title":"Physics-informed neural network (pinn) evolution and beyond: a systematic literature review and bibliometric analysis","volume":"6","author":"Lawal","year":"2022","journal-title":"Big Data Cognit. Comp"},{"key":"B31","doi-asserted-by":"publisher","first-page":"112291","DOI":"10.1016\/j.jcp.2023.112291","article-title":"Adaptive transfer learning for pinn","volume":"490","author":"Liu","year":"2023","journal-title":"J. Comput. Phys"},{"key":"B32","doi-asserted-by":"publisher","first-page":"208","DOI":"10.1137\/19M1274067","article-title":"Deepxde: A deep learning library for solving differential equations","volume":"63","author":"Lu","year":"2021","journal-title":"SIAM Rev"},{"key":"B33","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2103.03565","article-title":"Physics-aware deep neural networks for surrogate modeling of turbulent natural convection","author":"Lucor","year":"2021","journal-title":"arXiv"},{"key":"B34","doi-asserted-by":"crossref","DOI":"10.1007\/978-1-4899-7612-3","volume-title":"An Introduction to Mathematical Epidemiology","author":"Martcheva","year":"2015"},{"key":"B35","doi-asserted-by":"publisher","author":"Matthews","year":"2024","DOI":"10.48550\/arXiv.2408.10011"},{"key":"B36","doi-asserted-by":"publisher","first-page":"E1","DOI":"10.1038\/nature05523","article-title":"Hodgkin and huxley model\u2014still standing?","volume":"445","author":"McCormick","year":"2007","journal-title":"Nature"},{"key":"B37","doi-asserted-by":"publisher","first-page":"116172","DOI":"10.1016\/j.cma.2023.116172","article-title":"PINN-form: a new physics-informed neural network for reliability analysis with partial differential equation","volume":"414","author":"Meng","year":"2023","journal-title":"Comp. Methods in Appl. Mech. Eng"},{"key":"B38","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1109\/IJCNN60899.2024.10650694","article-title":"\u201cImportance of nyquist-shannon sampling in training of physics-informed neural networks,\u201d","volume-title":"2024 International Joint Conference on Neural Networks (IJCNN)","author":"Ooi","year":"2024"},{"key":"B39","doi-asserted-by":"publisher","first-page":"105583","DOI":"10.1016\/j.compfluid.2022.105583","article-title":"RPINNs: rectified-physics informed neural networks for solving stationary partial differential equations","volume":"245","author":"Peng","year":"2022","journal-title":"Comp. Fluids"},{"key":"B40","volume-title":"Differential Equations and Dynamical Systems","author":"Perko","year":"2013"},{"key":"B41","first-page":"1","article-title":"\u201cTorchdyn: implicit models and neural numerical methods in pytorch,\u201d","author":"Poli","year":"2021","journal-title":"Neural Information Processing Systems, Workshop on Physical Reasoning and Inductive Biases for the Real World"},{"key":"B42","doi-asserted-by":"publisher","first-page":"511","DOI":"10.1016\/j.jtbi.2009.07.040","article-title":"Comparing different ode modelling approaches for gene regulatory networks","volume":"261","author":"Polynikis","year":"2009","journal-title":"J. Theoret. Biol"},{"key":"B43","volume-title":"Mathematical Theory of Optimal Processes","author":"Pontryagin","year":"1987"},{"key":"B44","unstructured":"torchfunc\n          \n          2024"},{"key":"B45","doi-asserted-by":"publisher","first-page":"115980","DOI":"10.1016\/j.eswa.2021.115980","article-title":"Deep residual network for face sketch synthesis","volume":"190","author":"Radman","year":"2022","journal-title":"Expert Syst. Appl"},{"key":"B46","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2402.01868","article-title":"Challenges in training pinns: A loss landscape perspective","author":"Rathore","year":"2024","journal-title":"arXiv"},{"key":"B47","doi-asserted-by":"publisher","first-page":"109010","DOI":"10.1016\/j.cpc.2023.109010","article-title":"Seismicnet: Physics-informed neural networks for seismic wave modeling in semi-infinite domain","volume":"295","author":"Ren","year":"2024","journal-title":"Comp. Phys. Commun"},{"key":"B48","doi-asserted-by":"publisher","first-page":"469","DOI":"10.1177\/074873099129001037","article-title":"The goodwin oscillator: on the importance of degradation reactions in the circadian clock","volume":"14","author":"Ruoff","year":"1999","journal-title":"J. Biol. Rhythms"},{"key":"B49","first-page":"109","article-title":"\u201cIris recognition system using convolutional neural network,\u201d","volume-title":"2021 International Conference on Software Engineering and Computer Systems and 4th International Conference on Computational Science and Information Management (ICSECS-ICOCSIM)","author":"Sallam","year":"2021"},{"key":"B50","doi-asserted-by":"publisher","first-page":"10166","DOI":"10.1038\/s41598-023-36799-6","article-title":"Physics-informed neural ode (pinode): embedding physics into models using collocation points","volume":"13","author":"Sholokhov","year":"2023","journal-title":"Scientif. Reports"},{"key":"B51","doi-asserted-by":"publisher","first-page":"339","DOI":"10.1017\/S0956792519000081","article-title":"In search of a new economic model determined by logistic growth","volume":"31","author":"Smirnov","year":"2020","journal-title":"Eur. J. Appl. Mathem"},{"key":"B52","doi-asserted-by":"publisher","first-page":"1232","DOI":"10.1002\/2211-5463.12652","article-title":"Comparing deep belief networks with support vector machines for classifying gene expression data from complex disorders","volume":"9","author":"Smolander","year":"2019","journal-title":"FEBS Open Bio"},{"key":"B53","doi-asserted-by":"publisher","first-page":"674","DOI":"10.3390\/e25040674","article-title":"A second-order network structure based on gradient-enhanced physics-informed neural networks for solving parabolic partial differential equations","volume":"25","author":"Sun","year":"2023","journal-title":"Entropy"},{"key":"B54","unstructured":"Tripathi\n              S.\n            \n            \n              Farea\n              A.\n            \n            \n              Moroni\n              C.\n            \n            \n              Emmert-Streib\n              F.\n            \n          \n          Pinn Examples\n          \n          2024"},{"key":"B55","first-page":"147","article-title":"\u201cOn some important ordinary differential equations of dynamic economics,\u201d","author":"Tsoularis","year":"2021","journal-title":"Recent Developments in the Solution of Nonlinear Differential Equations"},{"key":"B56","doi-asserted-by":"publisher","first-page":"21","DOI":"10.1016\/S0025-5564(02)00096-2","article-title":"Analysis of logistic growth models","volume":"179","author":"Tsoularis","year":"2002","journal-title":"Mathem, Biosci"},{"key":"B57","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2205.13748","article-title":"Auto-PINN: understanding and optimizing physics-informed neural architecture","author":"Wang","year":"2022","journal-title":"arXiv"},{"key":"B58","doi-asserted-by":"publisher","first-page":"113112","DOI":"10.1016\/j.jcp.2024.113112","article-title":"A practical pinn framework for multi-scale problems with multi-magnitude loss terms","volume":"510","author":"Wang","year":"2024","journal-title":"J. Comput. Phys"},{"key":"B59","doi-asserted-by":"publisher","first-page":"60","DOI":"10.1016\/j.coisb.2021.03.005","article-title":"On structural and practical identifiability","volume":"25","author":"Wieland","year":"2021","journal-title":"Curr. Opini. Syst. Biol"},{"key":"B60","doi-asserted-by":"publisher","first-page":"18178","DOI":"10.1021\/acs.iecr.3c02383","article-title":"The application of physics-informed machine learning in multiphysics modeling in chemical engineering","volume":"62","author":"Wu","year":"2023","journal-title":"Indust. Eng. Chem. Res"},{"key":"B61","doi-asserted-by":"publisher","first-page":"15744","DOI":"10.1039\/D2CP05975D","article-title":"Analyzing drop coalescence in microfluidic devices with a deep learning generative model","volume":"25","author":"Zhu","year":"2023","journal-title":"Phys. Chem. Chem. Phys"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1717117\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,2,23]],"date-time":"2026-02-23T07:02:26Z","timestamp":1771830146000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1717117\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,2,23]]},"references-count":61,"alternative-id":["10.3389\/frai.2026.1717117"],"URL":"https:\/\/doi.org\/10.3389\/frai.2026.1717117","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,2,23]]},"article-number":"1717117"}}