Model Recovery at the Edge under Resource Constraints for Physical AI
This work addresses resource constraints for real-time, safe decision-making in mission-critical autonomous systems, representing an incremental improvement in edge deployment.
The paper tackles the inefficiency of deploying Model Recovery (MR) on edge devices for mission-critical autonomous systems by proposing MERINDA, an FPGA-accelerated framework that reduces DRAM usage by nearly 11x and speeds up runtime by 2.2x compared to mobile GPUs.
Model Recovery (MR) enables safe, explainable decision making in mission-critical autonomous systems (MCAS) by learning governing dynamical equations, but its deployment on edge devices is hindered by the iterative nature of neural ordinary differential equations (NODEs), which are inefficient on FPGAs. Memory and energy consumption are the main concerns when applying MR on edge devices for real-time operation. We propose MERINDA, a novel FPGA-accelerated MR framework that replaces iterative solvers with a parallelizable neural architecture equivalent to NODEs. MERINDA achieves nearly 11x lower DRAM usage and 2.2x faster runtime compared to mobile GPUs. Experiments reveal an inverse relationship between memory and energy at fixed accuracy, highlighting MERINDA's suitability for resource-constrained, real-time MCAS.