Lifeboat Foundation

For giant manufacturers, such as Germany-based Schaeffler Group with nearly 100 factories worldwide, 10 in the U.S., streamlining production of precision industrial components is paramount. When it began to digitize its operations in 2016 to get more real-time data from the shop floor, Schaeffler realized that it wasn’t enough to digitize the manufacturing processes, inventory, workforce, and logistics; it also had to digitize the physical manufacturing environment itself.

Its efforts to 3D scan each facility and create full-color lifelike digital twins, has, so far, resulted in significant time and cost savings. “Project lead times for data capture are up to 80% shorter compared to traditional methods,” says Roberto Henkel, Schaeffler’s senior VP of digitalization and operations IT. The full-color digital twins of its factories enables “far more efficient communication and planning among departments and third-party vendors.”

As the company accelerates its factory relocation to shorten supply chains and move production closer to customers, accurate factory digital twins have played a key role and also delivered some unexpected benefits.

The Geminid meteor shower is one of the most prolific annual meteor showers, impressing skywatchers year after year.

It is possible to see up to 120 meteors per hour under dark conditions when the Geminids peak each year.

Initially a variant of LSTM known as AWD LSTM was pre trained (unsupervised pre training) for language modelling task using wikipedia articles. In the next step the output layer was turned into a classifier and was fine tuned using various datasets from IMDB, yelp etc. When the model was tested on unseen data, sate of the art results were obtained. The paper further went on to claim that if a model was built using 10,000 rows from scratch then fine tuning the above model (transfer learning) would give much better results with 100 rows only. The only thing to keep in mind is they did not used a transformer in their architecture. This was because both these concepts were researched parallely (transformers and transfer learning) so researchers on both the sides had no idea of what work the other was doing. Transformers paper came in 2017 and ULMFit paper (transfer learning) came in early 2018.

Now architecture wise we had state of the art architecture i.e. Transformers and training wise we have a very beautiful and elegant concept of Transfer Learning. LLMs were the outcome of the combination of these 2 ideas.

A team of scientists from the United States, Italy, and China may have finally explained a large gap in the African and Eurasian fossil record. According to a model in a study published August 31 in the journal Science, the population of human ancestors crashed between 800,000 and 900,000 years ago. They estimate that there were only 1,280 breeding individuals alive during this transition between the early and middle Pleistocene. About 98.7 percent of the ancestral population was lost at the beginning of this ancestral bottleneck that lasted for roughly 117,000 years, according to the study.

[Related: Want more eye-opening science stories? Sign up for a PopSci newsletter.].

Humans have a need for serious speed—and this machine may be the answer.

A new gene therapy can reverse the effects of heart failure and restore heart function in a large animal model. The therapy increases the amount of blood the heart can pump and dramatically improves survival, in what a paper describing the results calls “an unprecedented recovery of cardiac function.”

Currently, heart failure is irreversible. In the absence of a heart transplant, most medical treatments aim to reduce the stress on the heart and slow the progression of the often-deadly disease. But if the gene therapy shows similar results in future clinical trials, it could help heal the hearts of the 1 in 4 people alive today who will eventually develop heart failure.

Researchers from City University of Hong Kong (CityUHK) and local collaborators have made a groundbreaking discovery of a new vortex electric field, poised to revolutionize future electronic, magnetic, and optical devices. This research holds immense promise for significantly enhancing the performance of various devices, particularly by improving memory stability and accelerating computing speeds.

Read Full Story.

You’ve probably heard of painful kidney stones. Find out where else “stones” can appear in your body, and how to tell.

Frequency-scanning systems with narrow instantaneous linewidth hold promise for various fields. Here, the authors report the use of time-variant parity-time symmetry to dynamically narrow the instantaneous linewidth of frequency-scanning systems.