Mechanical engineering articles from across Nature Portfolio

Pulse tube refrigerators are a critical enabling technology for many disciplines that require low temperatures, including quantum computing. Here, the authors show that dynamically optimizing the acoustic parameters of the refrigerator can improve conventional cooldown speeds up to 3.5 times.

Micro- and nanorobots present a promising approach for navigating within the body and eliminating biofilm infections. Their motion can be remotely controlled by external fields and tracked by clinical imaging. They can mechanically disrupt the biofilm matrix and kill the dormant bacterial cells synergistically, thereby improving the effectiveness of biofilm eradication.

Metal monochalcogenides — a class of van der Waals layered semiconductors — can exhibit ultrahigh plasticity. Investigation of the deformation mechanism reveals that on mechanical loading, these materials undergo local phase transitions that, coupled with the concurrent generation of a microcrack network, give rise to the ultrahigh plasticity.

An article in Nature Machine Intelligence presents an adaptable method to control magnetic microrobots’ navigation using reinforcement learning.

Minimally invasive surgery (MIS) lacks sufficient haptic feedback to the surgeon due to the length and flexibility of surgical tools. This haptic disconnect is exacerbated in robotic-MIS, which utilizes tele-operation to control surgical tools. Tactile sensation in MIS and robotic-MIS can be restored in a safe and conformable manner through soft sensors and soft haptic feedback devices.

3D printing can be used to automate the manufacturing of building elements for large-scale structures such as skyscrapers, aircraft, rockets and space bases without human intervention. However, challenges in materials, processes, printers and software control must first be overcome for large-scale 3D printing to be adopted for widespread applications.

Optimisation tasks in the inverse design of metamaterials with machine learning were limited due to the representations of generative models. Here the author comments a recent publication in Nature Communications which generates a latent space representation that unlocks non-linear optimisations.