Longer Wavelength Lasers for Inertial Fusion Energy with Laser-Plasma Instability Control: Machine Learning Optimum Spike Trains of Uneven Duration and Delay (STUD Pulses)

Default ARPA-E Project Image

OPEN 2021
Pleasanton, California
Project Term:
09/20/2022 - 03/19/2025

Technology Description:

Polymath Research will enable the use of longer-wavelength lasers for IFE. This project seeks to control LPI using pulses composed of Spike Trains of Uneven duration and Delay (STUD), a sequence of precisely timed laser pulses designed to disrupt LPI growth and memory build up in the plasma due to persistent self-organization of the plasma undergoing continuous and undisrupted laser energy deposition. The challenge is that with rather limited knowledge of the dynamic (micro-) state of the plasma, laser pulses composed of STUD must be devised to combat memory build up and exponential reamplification. The team will use data from simulation models and high-repetition-rate lasers to train a multitude of machine-learning algorithms to select optimal spike trains and define conditions where longer-wavelength, laser-triggered LPI can be successfully tamed. These predictions will then be tested in follow-on work on a laser facility operating at high energy.

Potential Impact:

Controlling LPI on the scale of instability growth times is a game changer for all laser-based IFE schemes.


The STUD-pulse program for LPI technology enhances the technical prospects of IFE for potentially abundant fusion power generation.


IFE is a promising non-carbon-emitting source of dispatchable primary energy.


The design and licensing of custom-made STUD pulses for LPI control for myriad IFE schemes using principled machine learning lends itself to potentially economical commercial fusion energy.


ARPA-E Program Director:
Dr. Ahmed Diallo
Project Contact:
Dr. Bedros Afeyan
Press and General Inquiries Email:
Project Contact Email:


Colorado State University

Related Projects

Release Date: