/Wraps/wrap09/controls/api/event_system/render has no handler for Purdue Event Documents of type 'Gilbreth Fellowship Research Proposal'

Quantum Acceleration of Sampling-Based Planning and Control

September 10, 2026

Share:

Project Description

Real-time decision-making in complex, uncertain environments is a fundamental challenge for autonomous systems, including ground, aerial, and space vehicles. Planning, control, and verification become particularly computationally demanding for nonlinear, stochastic, and high-dimensional systems. This research investigates how quantum algorithms can accelerate these critical decision-making tasks.

The proposed approach develops a unified pipeline that first reformulates planning, control, and verification problems as Monte Carlo sampling problems and then converts them into quantum estimation and search problems. Sampling-based methods are already widely used in autonomy and compatible with parallel architectures such as GPUs. For example, the path-integral controllers determine control inputs through Monte Carlo simulations of stochastic trajectories. These methods share a common computational bottleneck: large numbers of repeated model evaluations are required to estimate averages, probabilities, ratios, or extreme values.

Quantum algorithms offer a potentially transformative way to accelerate this bottleneck. Quantum amplitude estimation, for example, can provide a quadratic reduction in oracle queries for certain Monte Carlo tasks. We recently demonstrated this principle for a sampling-based controller. Building on this result, we study what sampling-based methods in planning, control, and verification can be converted into quantum query problems, and when quantum acceleration provides a meaningful end-to-end computational advantage.

Start Date

Spring/Summer/Fall 2027

Postdoc Qualifications

Successful candidates must hold a Ph.D. in mathematics, computer science, or related areas in engineering by the start date of the position, with a strong interest in quantum information science. Prior experience in one or more of the following areas is required: information science, operations research, control, autonomy, robotics, and machine learning.

Co-advisors

  • Takashi Tanaka, tanaka16@purdue.edu, AAE and ECE, URL: https://networked-control-systems-lab.github.io/
  • Vaneet Aggarwal, vaneet@purdue.edu, IE, ECE, and CS, URL: https://web.ics.purdue.edu/~vaneet/

Biliography

Y. Xu and V. Aggarwal, "Accelerating Quantum Reinforcement Learning with a Quantum Natural Policy Gradient Based Approach," in Proc. ICML, Jul 2025

D. Pedireddy, U. Priyam, and V. Aggarwal, "Noisy tensor ring approximation for computing gradients of variational quantum eigensolver for combinatorial optimization," Physical Review A, Vol. 108, Iss. 4, Oct 2023.

G. Das and T. Tanaka, “Model Predictive Path Integral Control as a Quantum Query Problem,” https://arxiv.org/abs/2607.28851, 2026

A. Patil, K. Morgenstein, L. Sentis, and T. Tanaka, Path Integral Methods for Synthesizing and Preventing Stealthy Attacks in Nonlinear Cyber-Physical Systems, https://arxiv.org/abs/2504.17118, 2025

Z. Wang, R. Keller, X. Deng, K. Hoshino, T. Tanaka, and Y. Nakahira, “Physics-Informed Representation and Learning: Control and Risk Quantification,” The 38th Annual AAAI Conference on Artificial Intelligence, 2024.