Informative, innovative and interesting articles from our favorite blogs
- Deep learning transforms smartphone microscopes into laboratory-grade devices, UCLA Henry Samueli School of Engineering of Applied Science, April 12, 2018, Science Daily -- Engineers have demonstrated that deep learning can discern and enhance microscopic details in photos taken by smartphones. This new technique improves the color details of smartphone images by so much that they resemble the quality of images from microscopes. This breakthrough could bring high-quality medical diagnostics into regions where people do not have access to high-end diagnostic technologies. The technique uses attachments that can be inexpensively produced with a 3D printer and cost less than $100 each. The team’s attachment can be placed over the smartphone lens to increase the resolution and visibility of tiny details in the images. The technique also uses artificial intelligence to reproduce the level of resolution and color details needed for a laboratory analysis. For the full story check out Science Daily.
- Space Technologies Helping Ecologists Track Endangered Animals, Michael Luciano, April 16, 2018, Wireless Design & Development -- Scientists are developing a system of drones and cameras that can record endangered species. Tracking elusive and endangered animals requires time, money, and resources, especially when the methods involve manual counts or taking photos from planes. The introduction of innovative technologies could ease some of these operational costs. The team is incorporating software originally used to find galaxies in space to identify the animals. One important commonality between animals and stars is they both emit heat and have recognizable thermal footprints. Currently, the scientists are teaming up with a safari park to film and photograph animals. They will need to gather thousands of images to form precise algorithms that identify target species in different ecosystems. In their first tests, the scientists targeted cows and humans and deployed their drones to test their algorithm’s ability to locate the cows in infrared footage. The team noticed that the algorithms mistook hot rocks for students posing as poachers. The tests did help the team to better calculate operational facets, like ideal heights to fly the drones. They hope to have a fully automatic prototype ready for testing in two years. For the full article visit Wireless Design & Development.
- Machine Learning Speeds Up Metallic Glass Discovery, Kenny Walter, April 16, 2018, Research & Development -- A new metallic glass is stronger and lighter than steel and can withstand corrosion and wear. A team of researchers used a new system at SLAC’s Stanford Synchrotron Radiation Lightsource (SSRL), which combines machine learning with experiments that quickly make and screen hundreds of sample materials at a time. The system allowed the team to discover three new blends of ingredients that form metallic glass 200 times faster than before. There are millions of possible combinations of ingredients for metallic class, but only a select few have been developed to where they can be commercially used. It can take up to two decades to get a material from discovery to commercial use. This new system gives scientists immediate feedback from machine learning models, allowing them to make and screen 20,000 samples a year. The new method could also be used to search for other materials and catalysts. For more information check out Research & Development.
- Researchers Find Combination for Small Data Storage and Tinier Computers, University of New Hampshire, April 13, 2018, Scientific Computing -- A computer the size of a pinhead may be a reality sooner than once thought. Researchers have discovered that using a combination of materials could be a way to offer a more stable environment for smaller and safer data storage. Current combinations of materials can create volatile situations in which data can be lost once the device is turned off. Researchers have found that their new combination is a much safer option. The team’s proposed method allows for a stable perpendicular anisotropic energy (PMA), which is the key component in a computer’s data storage. The team’s unique combination would allow for larger amounts of data to be stored in smaller environments. The researcher’s work opens the door to possibilities for much smaller computers for everything from basic data storage to space missions. For the full article visit Scientific Computing.
- And lastly, the most popular Tektronix download of the week goes to – Data Logging of Power Profiles from Wireless IoT and Other Low-Power Devices. Our latest application note illustrates different methodologies that can be used for capturing the load current characteristics of a device or the voltage discharge profile of a battery.
Download your copy today!