# Enabling Physical AI at the Edge: Hardware-Accelerated Recovery of System Dynamics

arXiv: https://arxiv.org/abs/2512.23767 · HTML: https://mlsanity.com/paper/2512.23767/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 14 of its 31 references that are in the map.
- First posted on arXiv 29 Dec 2025.

## Builds on (most cited first, 14 of 14)

- [From Words to Watts: Benchmarking the Energy Costs of Large Language Model Inference](https://mlsanity.com/paper/2310.03003/) (2023)
- [Physics-informed learning of governing equations from scarce data](https://mlsanity.com/paper/2005.03448/) (2020)
- [Liquid Time-constant Networks](https://mlsanity.com/paper/2006.04439/) (2020)
- [Sparse identification of nonlinear dynamics for model predictive control in the low-data limit](https://mlsanity.com/paper/1711.05501/) (2017)
- [Neural Flows: Efficient Alternative to Neural ODEs](https://mlsanity.com/paper/2110.13040/) (2021)
- [Energy Considerations of Large Language Model Inference and Efficiency Optimizations](https://mlsanity.com/paper/2504.17674/) (2025)
- [Apple Intelligence Foundation Language Models: Tech Report 2025](https://mlsanity.com/paper/2507.13575/) (2025)
- [Edge AI: A Taxonomy, Systematic Review and Future Directions](https://mlsanity.com/paper/2407.04053/) (2024)
- [Physics guided neural networks for modelling of non-linear dynamics](https://mlsanity.com/paper/2205.06858/) (2022)
- [Recovering implicit physics model under real-world constraints](https://mlsanity.com/paper/2412.02215/) (2024)
- [Accelerating ODE-Based Neural Networks on Low-Cost FPGAs](https://mlsanity.com/paper/2012.15465/) (2020)
- [Model Recovery at the Edge under Resource Constraints for Physical AI](https://mlsanity.com/paper/2512.02283/) (2025)
- [Fast Online Digital Twinning on FPGA for Mission Critical Applications](https://mlsanity.com/paper/2512.17942/) (2025)
- [Accelerated Digital Twin Learning for Edge AI: A Comparison of FPGA and Mobile GPU](https://mlsanity.com/paper/2512.17941/) (2025)

## Abstract (arXiv)

Physical AI at the edge -- enabling autonomous systems to understand and predict real-world dynamics in real time -- requires hardware-efficient learning and inference. Model recovery (MR), which identifies governing equations from sensor data, is a key primitive for safe and explainable monitoring in mission-critical autonomous systems operating under strict latency, compute, and power constraints. However, state-of-the-art MR methods (e.g., EMILY and PINN+SR) rely on Neural ODE formulations that require iterative solvers and are difficult to accelerate efficiently on edge hardware. We present MERINDA (Model Recovery in Reconfigurable Dynamic Architecture), an FPGA-accelerated MR framework designed to make physical AI practical on resource-constrained devices. MERINDA replaces expensive Neural ODE components with a hardware-friendly formulation that combines (i) GRU-based discretized dynamics, (ii) dense inverse-ODE layers, (iii) sparsity-driven dropout, and (iv) lightweight ODE solvers. The resulting computation is structured for streaming parallelism, enabling critical kernels to be fully parallelized on the FPGA. Across four benchmark nonlinear dynamical systems, MERINDA delivers substantial gains over GPU implementations: 114× lower energy (434~J vs. 49,375~J), 28× smaller memory footprint (214~MB vs. 6,118~MB), and 1.68× faster training, while matching state-of-the-art model-recovery accuracy. These results demonstrate that MERINDA can bring accurate, explainable MR to the edge for real-time monitoring of autonomous systems.