I will report on two recent works in the He* BEC lab. The first on the presence of Bogoliubov-Cherenkov Radiation in an Atom Laser(pictured) and the second on the time-optimal control of BEC’s using machine learning.
Hypersonic flows are well studied in classical gases, however, for quantum fluids the classical tools such as shock tunnels are unavailable. We develop a simple yet powerful technique to probe the high-hypersonic regime in quantum gases by using an atom laser out-coupled from a tightly confined Bose-Einstein Condensate. The atom laser expands rapidly to form a dilute gas with a low speed of sound and as it falls encounters the original condensate as a strong repulsive impurity. As the passage is much greater than the local speed of sound (M~50) we observe the quantum fluid analog of Cherenkov radiation: Bogoliubov-Cherenkov radiation which produces a striking fringe pattern as observed in the far field using a single atom sensitive delay line detector (pictured). Our work provides a controlled system for studying analogs of exotic phenomena such as the dynamic Casimir force, quantum friction, the Unruh effect and Hawking radiation.
Fast transport of quantum states is a vital part of developing quantum technologies, where manipulations must race against decoherence. Traditional approaches either use adiabatic methods that take prohibitively long times to execute or “shortcut” schemes that rely on a precise theoretical model of the system to produce reasonable performance. For a test case, we employ the difficult high dimensional problem of transporting a BEC between two vastly different traps. In this system we present a theory-free method based on an empirical machine learning algorithm to produce optimal transport close to the adiabatic limit, outperforming both previous approaches. This work provides a general method for the future control of complex high dimensional systems such as quantum computers and sources.