Ising Machines for Model Predictive Path Integral-Based Optimal Control

NeurIPS 2025 Workshop FPI Submission58 Authors

Track: Main Track

Keywords: model predictive control, p-bits, ising machines, optimal control, model predictive path integral, probabilistic computing, sampling, boltzmann, gibbs sampling, qubo

TL;DR: We make use of Ising Machines to sample a Boltzmann distribution so that probabilistic p-bit computers can perform Model Predictive Path Integral Optimal Control

Abstract: We present a sampling-based Model Predictive Control (MPC) method that implements Model Predictive Path Integral (MPPI) as an Ising machine, suitable for novel forms of probabilistic computing. By expressing the control problem as a Quadratic Unconstrained Binary Optimization (QUBO) problem, we map MPC onto an energy landscape suitable for Gibbs sampling from an Ising model. This formulation enables efficient exploration of (near-)optimal control trajectories. We demonstrate that the approach achieves accurate trajectory tracking compared to a reference MPPI implementation, highlighting the potential of Ising-based MPPI for real-time control in robotics and autonomous systems.

Submission Number: 58