The primary failure in mechanical elements is gear contact fatigue — but it’s not clear how exactly these failures occur, according to an international team of researchers. They recently reviewed the latest developments in the field to frame out what’s known and the remaining questions for further work. In addition, they conducted their own gear fatigue assessments with multiple materials and production methods, from which they have proposed formulas for others to use in their evaluations.

Every mechanical assembly produces acoustic emissions. Cranks, bearings, switches and far more make noise — the kind that can be heard and the kind that advanced technology can detect to determine the health of the mechanism. One area where this acoustic emission technology has not kept up, though, is for journal bearings: the support for rotating shafts in various machines, including transmission engines. Journal bearings depend on a film of oil surrounding the shaft to maintain operation. Should the oil become contaminated or some other mechanical failure occur, the entire mechanism can be at risk. Researchers have studied acoustic emissions for detecting mechanical failure since the 1950s and specifically for journal bearings since 1985. Now, an international team has pulled together the latest in the field to better understand how far it has come and what direction it should go.

Photodetectors require complex evaporation-based fabrication processes and exhibit relatively high dark currents and noise levels under operation. Besides, system size, cost, and the mismatch in responsivity between different types of photodetectors hinder further commercial application. Now, researchers from Dalian Minzu University have proposed the construction of a perovskite/In0.53Ga0.47As thin-film heterojunction, thereby realizing broadband photodetectors covering the deep-ultraviolet to short-wavelength infrared range.

Inspired by fish lateral lines, researchers developed a lightweight, miniaturized sensing unit that estimates both flow speed and direction. Combined with a neural network and an inertial measurement unit, the single unit was validated on an underwater robot during multiple motion states and attitude disturbances, reflecting practical underwater operating conditions.

Seed-grown Cannabis sativa crops often vary in flowering time, yield, sex expression, and chemical profile, making large-scale production less predictable.