We discuss the design, operation, and performance of a vacuum setup constructed for use in zero (or reduced) gravity conditions to initiate collisions of fragile millimeter-sized particles at low velocity and temperature. Such particles are typically found in many astronomical settings and in regions of planet formation. The instrument has participated in four parabolic flight campaigns to date, operating for a total of 2.4 h in reduced-gravity conditions and successfully recording over 300 separate collisions of loosely packed dust aggregates and ice samples. The imparted particle velocities achieved range from 0.03 to 0.28 m s(-1) and a high-speed, high-resolution camera captures the events at 107 frames/s from two viewing angles separated by either 48.8 degrees or 60.0 degrees. The particles can be stored inside the experiment vacuum chamber at temperatures of 80-300 K for several uninterrupted hours using a built-in thermal accumulation system. The copper structure allows cooling down to cryogenic temperatures before commencement of the experiments. Throughout the parabolic flight campaigns, add-ons and modifications have been made, illustrating the instrument flexibility in the study of small particle collisions.

Machine learning (ML) is an effective tool to interrogate complex systems to find optimal parameters more efficiently than through manual methods. This efficiency is particularly important for systems with complex dynamics between multiple parameters and a subsequent high number of parameter configurations, where an exhaustive optimisation search would be impractical. Here we present a number of automated machine learning strategies utilised for optimisation of a single-beam caesium (Cs) spin exchange relaxation free (SERF) optically pumped magnetometer (OPM). The sensitivity of the OPM (T/Hz), is optimised through direct measurement of the noise floor, and indirectly through measurement of the on-resonance demodulated gradient (mV/nT) of the zero-field resonance. Both methods provide a viable strategy for the optimisation of sensitivity through effective control of the OPM’s operational parameters. Ultimately, this machine learning approach increased the optimal sensitivity from 500 fT/Hz to <109fT/Hz. The flexibility and efficiency of the ML approaches can be utilised to benchmark SERF OPM sensor hardware improvements, such as cell geometry, alkali species and sensor topologies.