Natural guar-based active ingredient extends the range of silicone-free beauty care solutions for stress-free hair styling Solvay, a leader in naturally derived ingredients for beauty care formulations, introduces Polycare ® Heat Therapy, a new functional ...

Solvay launches Polycare Heat Therapy for bio-based thermoprotection in hair care

Get the most recent info and news about Polycare on HackerNoon, where 10k+ technologists publish stories for 4M+ monthly readers.

To drive empowerment and circular economy, Polycare develops innovative construction technology, that makes sustainable habitats affordable.

Polycare

Introduction, concerning the time when the Apocalypse was written

PLUTONIC AND METAMORPHIC ROCKS:—CLEAVAGE AND FOLIATION

GM. It's Crypto Player One, the first and only daily newsletter on crypto gaming. We are the Drake of web3 gaming, consistently giving you the hits. Here's what we have for you today: Daily Trivia. Sony is dipping its toes in blockchain tech. In Other News. Daily Trivia Who are the original designers of the Mortal Kombat franchise? Scroll down to find out the answer. Sony Files a Blockchain and NFT-Related Patent Sony has a new patent that will allow them to track changes in ownership, metadata, and visual appearance of digital assets. These assets could be moments of gameplay, in-game characters, or items. According to the patent, Sony can track the history of the in-game items and see all its previous owners. They want to use a blockchain to record these items and include a token with a unique identifier. Last year, PlayStation (owned by Sony) filed this patent before gamers started losing their minds about gaming NFTs. Still, Sony and PlayStation have played around with NFTs in the last few months. During the EVO 2022 event, Sony asked fans what kinds of NFTs they would want to buy. Also, Sony partnered with Theta Labs to launch an NFT collection. PlayStation has its own Stars rewards program that features digital collectibles. The company has said that these items "are not NFTs" because you can't trade them. But with the new patent, it looks like Sony plans to explore and experiment on the blockchain. This is ultra bullish for the blockchain gaming space because Sony is one of the biggest game publishers. In Other News L’Oréal partners with the leading cross-game avatar platform, Ready Player Me. Polycade and Atari team up to create Web3 games and art collections. The layer 1 blockchain Aptos will partner with a South Korean gaming studio to focus on web3 gaming. Answer Ed Boon and John Tobias. New to the Newsletter? Sign up to get the first and only daily newsletter on web3 gaming — it’s free! Photo by on Nikita Kostrykin Unsplash

🎮 Sony Potentially Entering the World of Blockchain Gaming

Over 143 million Americans, or 64% of U.S adults currently wear prescription eyewear, it’s something that takes time, money, and attention to stay up to date on. Costing an average of $173 per pair of glasses and needing to be replaced every 1-3 years, here’s a few hacks to make that process a bit easier and to find more suitable glasses. It’s important to first look at all the options available when getting a new pair of glasses. Most people are going to get simple single vision lenses, but that’s far from the only option. Bifocals and trifocals have risen in popularity as a good option to see at distinct different distances. Although progressive lenses do the same but with a more gradual change in the lenses prescription. This is a good way to get around the clunky look and feel of bi/trifocals without losing out on their benefits. High-index lenses are another modern lens type that make stronger prescriptions more wearable. There are also the three versions of lens material, plastic, polycarbonate, and glass, each with a specific set of positives and negatives.  Glass being the least durable but most clear, polycarbonate being quite durable but a bit more expensive, and plastic being a good middle ground between the two. Tinted lenses have also been growing as one of the most unique lens features available. No longer are lens tints just an aesthetic option, but they can be really helpful to those with select issues. Pink lenses, for example, can reduce migraines and help with depth perception. Other effects for different colors include fatigue decrease, strain decrease, reflection protection, and a contrast increase. These are all great options to have, but can definitely get a bit pricey. For those buying a new pair of glasses every few years, it’s going to be a hard sell to ask for an extra $100 in new lens features. This is where replacement lenses come into play. work by replacing the lenses in a pair of glasses alone, not requiring the additional purchase of the frames as well. This makes getting a new pair of glasses easier, quicker, and cheaper. Replacement lenses It also allows a greater sense of individuality in one's glasses. Vintage and more unique styles found outside of a retailer can be sent in to be filled with modern lenses. No longer are the walls of frames at retailers the only option available. Finally reusing the same frames over several years allows more lens options to be used with the money saved on frames. Instead of buying a new pair of frames every 1-3 years, the consumer can now opt to tint their glasses, get a higher quality pair of lenses, etc. The main hack that must be realized when buying a new pair of glasses, there are options. There’s countless options in frames, lenses, where to buy glasses, the price of glasses, there’s something available for everyone in every price range. The important thing is to consider every option and get the best pair for each individual.

Replacement Lenses: Hacking the Need for New Glasses

The pandemic with its subsequent lockdowns is forcing artists to search for innovative ways to display and sell their work. And while many people welcome technologies, such as artificial intelligence (AI) and the Internet of things (IoT) in art distribution, they are skeptical about its ability to generate art. They think such pieces will lack creativity and will generally be out of place. Here is one example to challenge this belief. Marcus du Sautoy, a professor of mathematics at Oxford University and an author of several books, confessed that he for one of his books. The exciting thing is that no one noticed these words were artificial. Not even his editor. used an algorithm to generate 350 words This article elaborates on how combining technology and arts will bring innovative ways to artwork creation, display, protection, and monetization. Maybe it will inspire you to benefit from to bring your ideas to life. artificial intelligence as a service The History of Digital Art and Technology The first attempt to merge visual arts and technology dates back to the ’60s when New York-based artists teamed up with scientists and engineers from Bell Laboratories to . They used innovative, at that time, wireless sound transmission and video projection technologies. create installations and deliver remarkable performances When speaking of influential artists mixing tech and arts, Andy Warhol immediately comes to mind. He was among the first to create digital drawings on an Amiga computer in 1985. Unfortunately, the piece was lost and by one of Warhol’s fans. Since then, modern technology has continued to influence the way artists create, display, and sell their work. recovered only in 2014 When COVID-19 hit, the arts sector struggled immensely. Americans for the Arts surveyed art workers to estimate the impact of the pandemic on this industry. The study revealed while 24% of organizations were forced to reduce staff. And yet, 65% of the respondents are still managing to deliver content. Let’s see how technology impacts the arts sector and helps it overcome the pandemic-induced challenges. that 95% of art-related events were canceled How is the New Technology Used to Create Art? AI and Art Creation Christie’s art auction company recently sold its . The painting is a blurred face titled Portrait of Edmond Belamy. Generative adversarial networks (GANs) are typically used to produce this type of distorted and weird imagery. At first glance, such work might look odd, but some human artists created similar art on purpose. Think of Francis Bacon’s deformed portraits, such as Three Studies of a Portrait of Henrietta Moraes. first AI-generated art piece for $432,500 Source: Americanscientist.org Artificial intelligence-generated art can also closely resemble human creations. One example is the tool, which was trained on 100,000 paintings of celebrated artists, such as Bruegel and Rembrandt. There are plenty of other . AICAN creative AI-powered apps to help artists generate paintings To understand how individual artists deploy GANs to produce images and video, we can look at Amir Zhussupov’s work. This artist from Kazakhstan selects different datasets based on his interest to train a GAN. Amir relies on code, but there is still plenty of work left to do by hand. Here is what Amir told STIR, “For the (GAN) model I have been working on for the past month, I had to clean up, color correct, crop and align more than 17,000 paintings. So I have been practically living inside Photoshop, ha-ha!” The piece below is Amir’s GAN-based work inspired by famous science fiction and horror artist HR Giger. Another example of the art and AI powerful combo comes from Refik Anadol. The artist used 500,000 images, several thousands of audio recordings from the LA Philharmonic’s archive, and hundreds of videos to train a series of algorithms to combine all this data and deliver images that he later projected on the Walt Disney Concert Hall for ten days. Here is one more example from ITRex. A renowned artist contacted us to build an AI-powered tool that would generate paintings based on the artist’s existing work together with paintings of other artists who inspired our client. ITRex team used neural networks to learn the artists’ trademarks and generate paintings accordingly. Robots in Art Creation Examples of robots in the art industry vary in their complexity, from mini robots that draw colorful lines following clear instructions to self-sufficient robot artists that can imitate humans in creating art. Artists Julian Adenauer and Michael Haas created a robot that moves around on a canvas leaving a colorful trail. This art piece is installed on the wall of the Berlin Gallery. It is changing every day as the robot keeps moving around, adding color in response to , “the process of creation is ideally endless.” Haas’s idea of creation Moving on to a more complex example of robotic technology in arts, Ai-Da, a robot artist, has human features and a robotic arm system. This AI-powered robot can analyze the images in front of her and command her arm movement to produce sketches. Ai-Da solely dedicated to an AI artist. participated in the world’s first exhibition Source: Time.com Sometimes, artists use robots to assist in art creation instead of relying on them to work autonomously. Artist Sougwen Chung designed and developed a robotic arm that helps her with painting. It connects to Chung’s brain-wave data, which influences the way the robot behaves. Chung believes this is a true human-machine partnership, and she is inspired to co-create together. In the pre-pandemic era, Chung performed with her robot in front of a live audience in a gallery setting. She was . Now, she streams her collaborations live into exhibit spaces from her studio. selling the resulting paintings for over $131,000 apiece How AR and VR Help Artists The gaming and entertainment industries were the first beneficiaries of extended reality. Soon, individual artists, museums, and galleries also turned to VR and . Among early applications of augmented reality in arts is the work of a Japanese artist, Nubumichi Asai. He gathered a team of technology experts and makeup artists to create an application that projects patterns on a moving face. This software can detect the human face and project strictly within its contours. AR development companies Another exciting virtual reality application in art is , which allows artists to immerse into their own creation and view it from different angles. They can draw large objects floating in the air without paying attention to gravity. Google’s Tilt Brush Speaking of recent VR and AR applications in the arts domain, an extended reality art production company, , collaborated with Olafur Eliasson to . Eliasson specializes in resembling natural phenomena and the outdoors in his pieces. Now, with the help of augmented reality, users can project northern lights, sun, rainbow, clouds, or any other natural manifestations into their living room. Acute Art bring his artwork into art lovers’ personal space Sensors and IoT in Art Creation This technology enables artists to gather data from various locations and project it into their artwork. One example comes from San Jose International Airport that . This artwork resembles a cloud whose polycarbonate tiles change from transparent to opaque state as it reads real-time weather data. hosts an eCLOUD A Russian artist, Dmitry Morozov, offers another interesting example of using sensor technology in arts. Morozov created a Bluetooth-connected portable tool that captures pollution data and translates it into sketches. The device is based on sensors that measure dust and different gazes, such as carbon dioxide, carbon monoxide, and methane and translate this data into volts, which are algorithmically manifested in different shapes and colors on a canvas. Morozov thinks back to the moment he got inspired to create this tool, “I wondered, what can I do with sensors, a small printer and the polluted air in Moscow?” Now, thanks to his artwork, people can see which city areas are more polluted than others. Ironically, the brighter the image, the more polluted the air. Source: Wired.com Technology Changes the Way Art is Displayed Virtual Museum Tours Due to the pandemic, many countries went on lockdown, and visiting museums physically became impossible. With modern technology in arts, you can see exhibitions while sitting on your sofa. Google Arts and Culture collaborated with over to allow art lovers to immerse themselves in the art world online through virtual tours. Now one can view the history of American fashion in the time of 1740 to 1895 at the National Gallery of Art in Washington. Then move to Paris to enjoy the works of famous French artists at Musée d’Orsay. Then virtually travel to the Netherlands and witness incredible art pieces from the Dutch Golden Age, including the works from Rembrandt and Vermeer at the Rijksmuseum in Amsterdam. 2,500 galleries and museums worldwide Live Streams This technology is widely . Now, live streams of exhibitions and concerts are also gaining popularity due to COVID. Towards the end of 2020, a famous English singer Dua Lipa streamed her concert live, and it was a vast success. According to the singer’s team, around watched the event in real time. China alone brought in almost two million attendants. Like concerts, exhibitions can also be streamed with the help of new art technologies. One example comes from Canada. Three Canadian galleries featuring one famous painting, Wayne Gretzky #99. One gallery holds the original painting while the other two display replicas. This live stream captured the artwork and the viewers observing it. So, when visitors arrive at The Room, the gallery with the original painting, they are captured on a video together with the painting and streamed to the other two galleries. This way, the live stream did not only feature the painting, but also the people who were interacting with each other and goofing around. used in the retail sector 5 million viewers from different countries teamed up with a tech partner to organize a live stream IoT Systems In addition to creating art, can help museum staff protect paintings while improving visitor experience. Common IoT applications in the arts domain include: IoT technology solutions Preserving artwork: paintings are sensitive to humidity and light. Prolonged exposure to moisture and direct sunlight will cause decay and discoloration, among other problems. Thanks to technology, museums can protect their art pieces by installing IoT sensors to measure temperature, light, and other relevant factors in real time. These sensors’ readings allow museum staff to adjust the relevant parameters to preserve the paintings and minimize restoration projects. Another benefit of IoT technology in arts is protecting installations from water leakage. Museums can place sensors on water pipes near display areas to signal any leaks, enabling the staff to address the problem before it causes damage. Supplying museum workers with helpful stats: IoT sensors can give museum staff an indication of which art pieces are the most popular among visitors. Sensors can track presence density around installations in real time to spot the artworks in demand. Another exciting application of sensors is measuring visitors’ heart rate and respiration to detect their emotional response to paintings and determine which ones elicit the most reaction. Improving navigation: beacons positioned at strategic locations can communicate with the visitors’ smartphone sending notifications and eliciting action. For example, if a visitor downloaded the museum’s app, beacons could trigger the app to tell a story about a particular painting. Such sensors can also improve navigation by telling visitors where their favorite painting or art object is located. Technology Facilitates Art Trade Online Platforms Online auctioning platforms put one more marketing tool at the artist’s disposal. A seemingly easy option for beginner artists would be to sell their work on the . However, this platform has its seller approval process. Artists will need to meet specific production volumes that might be larger than intended. On top of that, their work might be placed outside the Arts category, making it difficult for customers to find. Artists who have already gained recognition can sell their artwork on more professional and specialized mediums, such as and art auctions. A creative technology studio turned to ITRex to build an . This solution would allow artists to promote their work by decorating virtual rooms with their art pieces. Fans would be able to visit their favorite artists’ rooms and share their experiences with friends. Fans could also buy selected items in a non-fungible token (NFT) marketplace. dedicated section on Amazon Christie’s Artnet AR-based marketing app for fan engagement Blockchain for Art Tokenization and Crypto Trading Blockchain is another innovative technology used in the arts sector. One example is trading crypto collectibles. platform offers 10,000 unique collectible characters. Anyone on the Ethereum blockchain can own a character. When the project started, people could obtain a character for free and have ownership proof. All they needed was an Ethereum wallet. All the 10,000 punks were quickly claimed. Now, users can place bids and sell characters following the instructions on the platform. Here is a more innovative approach to selling art. A London-based startup, Maecenas, to the arts industry in 2018. This technology allows artists to divide their artwork into fractions. Every fraction can be sold and owned separately. One buyer can purchase several fractions of one painting or even fractions of several paintings, creating a diverse portfolio. In 2018, Maecenas marketed the 14 Small Electric Chairs painting by Andy Warhol. At that time, it was valued at $1.7 million, but the blockchain-based sale generated excessive attention, skyrocketing the painting’s value to $5.6 million. CryptoPunks introduced the concept of tokenization Technology Improves Artwork Security Statistics on art theft indicate that of artwork are stolen each year. Luckily, technology can help the art segment here too. AI and IoT sensors can recognize unauthorized presence on a museum’s ground and detect artwork forgery, among other fascinating applications. over 50,000 pieces Detecting forgery: there is evidence of AI’s potential to see a forgery in paintings. Steven and Andrea Frank on artwork belonging to an artist and then run the algorithm on a painting whose ownership is debatable to see if it matched that artist’s style. One of the main problems the couple faced was a lack of training material. They could overcome this by generating tiles (smaller fragments of a painting) from available artwork. Using this algorithm, the couple could correctly attribute paintings to Rembrandt, van Gogh, and other artists. were able to train convolutional neural networks (CNN) Access control: museums can place IoT sensors on doors and windows to give a signal, silent or loud, enabling security guards to react to intrusions. Also, sensors can be placed around valuable artwork, sending alerts if someone touches it. Occupancy sensing: presence identification sensors can be positioned to detect movement when the museum is closed for visitors. Technology and Art Restoration and Reconstruction AI-based artistic technology can help restore paintings. Just recently, Rijksmuseum in the Netherlands . This artwork, from 1642, was moved to Amsterdam’s City Hall where officials decided to cut large panels from the sides of the painting to fit it in a location they fancied. In 2019, Rijksmuseum dedicated millions of dollars to building and training of an AI tool that would reconstruct the missing parts in their original style. When asked why the museum went with AI instead of hiring an artist to complete the piece, Rob Erdmann, a senior scientist heading this project answered, “There’s nothing wrong with having an artist recreate [the missing pieces] by looking at the small copy, but then we’d see the hand of the artist there. Instead, we wanted to see if we could do this without the hand of an artist. That meant turning to artificial intelligence.” In another fascinating example, AI could . When the artist was short on money, he tended to draw on top of his work. Conservators studying The Crouching Beggar noticed there was another image underneath. The Art Gallery of Ontario released an X-ray of this forgotten artwork enabling scientists to use AI and 3D printing to reconstruct the piece. used AI to restore The Night Watch by Rembrandt recreate a landscape hidden under one of Picasso’s paintings Art and Technology Relationship: What to Expect in the Future Some people are still skeptical about the future of new technology and arts. A renowned art critic and Pulitzer Prize winner, Jerry Saltz, Despite some people’s pessimism, this article presents many applications of modern technology in the arts sector. It can support artists at different stages of their project's lifecycle. And when it comes to AI, it doesn’t have the same creativity as humans (yet), but it has its unique usage. For example, after proper training, AI can imitate a style of a renowned painter from the past without adding its own touch, as humans would do. There are issues that we need to address when resorting to technology in the arts. For example, in online streaming, paying the intellectual property royalties became a complicated matter. Some live stream concerts have when they understand how royalties work in this context. Standard licensing frameworks are also still missing. Another concern is who owns the copyrights of AI-generated art. Despite these obstacles, technology widens our horizons whether we produce or simply enjoy arts, and all the examples mentioned in this article solidify the relationship between art and technology. If you have an idea of enhancing your studio, museum, or art research with technology, don’t hesitate to reach out. believes that AI-generated art is “boring and dull.” allocated money to be paid Whether you are concerned with enhancing art creation, display, or monetization, ! Our experts will put modern technology at your service. drop ITRex a line

Technology has Brought New Innovations to How Art is Made, Displayed, and Sold

Materials and equipment needed To build and test these test lungs, some equipment is essential. The following is a list of suggested gear: Compressed air source: oil free compressor, several cfm at > 50psi Recommended: Porter-Cable C2002 Pressure gauge: should read 0–50 cmH2O range with <1 cmH2O resolution, preferably with differential inputs, preferably digital Recommended: Risepro HT-1890 Flow meter: should read 0–60 lpm with < 1 lpm resolution, preferably digital Recommended: no recommendation yet Rigid vessel: one 14 gallon or three 5 or 6 gallon glass carboys (from home brewing store) (or) Flexible bag: anaesthesia bag, 2–3 liters (or other similar size bag) Recommended: Smiths Portex #370924 Plastic sheets (for springs): 12x12 inch, 1/8 thick, HDPE, polypropylene, or polycarbonate Recommended: McMaster 2898K11 https://www.mcmaster.com/2898k11 Tubing: 3/4 inch PVC pipe, schedule 40 (from hardware store) — By Alex Izvorski and Robert L. Read Teams building and testing ventilators for COVID-19 need to have “test lungs”. The term “test lung” sometimes means a physical air-tight container with sophisticated instrumentation added, but here we mean just the physical container without the instruments. A test lung is a simulator of a real pair of lungs, but made in such a way that it has standard and stable physical properties. In particular, it should press against air inside it when inflated, just like real lungs, which is called . It should oppose air from moving in or out too fast (just like real lungs which can’t fill or empty instantly) this is called . It is almost impossible to understand how an emergency pandemic ventilator designed to help humans breathe will behave without testing on test lungs. Engineers building prototypes need a way to test on their workbench, even if much more rigorous testing will eventually be needed before a ventilator is used on a patient. This article teaches how to model compliance and resistance, and will be useful if you need to understand those concepts, or if you actually need to test a pandemic ventilator. compliance resistance (A Commercial test lung with a resistance tube and barbed bleed adapter.) How Doctors Measure Pressure For respiration, only pressure differences matter. Breathing is driven not by atmospheric pressure, but by difference in pressure in the airway compared to atmospheric pressure. Although engineers normally use the SI unit of pressure, which is the Pascal, the medical profession usually measures pressure in terms of the height of a column of water it will raise — -hence, centimeters of water, denoted as cmH2O. 1 cmH2O is approximately equal to 98 Pa. It takes 70 cm H2O to make one psi. Breathing pressure differentials almost never go above one psi! To measure pressure inside an airway, you have to get a pressure meter probe or sensor sealed inside the airway. Adult airways use an international 22mm standard. A barbed bleed adapter is a convenient way to insert a pressure sensor into an airway. This is about the size of ¾” PVC tubing, which is widely available. (A smaller port in the airway makes taking a pressure measurement easier.) Compliance Typical healthy lungs have a compliance around 20–50 milliliters/cmH2O. In other words, pushing in 20–50 ml air would raise pressure by 1 cmH2O. (Different lungs or diseases have different compliance curves.) The compliance of real lungs is not completely constant. Just like a balloon getting tight as it gets full, the lungs require more pressure to push or pull another milliliter in as they get more full. That is why a plot of pressure vs. volume is a typical S-shaped curve. Some disease conditions make the lungs stiff (low compliance), or even loose (high compliance). Because applying too much pressure to a patient’s lungs or blowing too much air into them causes serious injury, ventilators have to be tested against realistic compliance and resistance. Resistance Typical healthy lungs also have a resistance around 3–7 cmH2O/(liters/second). Thus, pushing air in at a constant flow rate of 1 liter/second (60 liters per minute) would raise pressure by 3–7 cmH2O (in addition to the change due to compliance). A realistic test lung is simply a combination of an element with a known resistance and an element with a known compliance. Commercial Test Lungs Instrumented commercial test lungs are available, but may be quite expensive, and may not be easy to obtain in the current crisis. Some examples of instrumented test lungs are the Michigan Instruments Model 1600 Test Lung; the IngMar ASL 5000 Breathing Simulator; and the IMT SmartLung. Even uninstrumented simple plastic test lungs may be hard to obtain in some places during the pandemic. If you are working on a ventilator project and have one of these, great! If you don’t, this article is for you. Constructing Your Own Test Lungs Several ways of constructing test lungs with known resistance and compliance are described in the ISO 10651–5 standard for ventilators. It turns out that some of those don’t require any expensive or difficult to obtain materials or equipment. They work as well as the commercial test lungs, and are suitable for certifying designs. To calibrate your test lung, you will need a pressure meter or sensor and a way to take a measurement just inside the throat of the test lung, such as by using a CPAP bleed adapter or some other Y-connector or T-connector that you can seal around your pressure meter. Constructing Compliance from a Glass Vessel Compliance is measured in milliliters/cmH2O. This may be concisely written as just a C-value without the units, and those units are assumed. The typical adult range is around C20 (+-10), mainly depending on body size and physical condition. Compliance may decrease greatly with conditions that cause the lung tissue to stiffen (eg ARDS). Values used in the ISO standards (for adults) are: C50 — large adult C20 — normal C10 — unhealthy lungs : a rigid vessel such as a large glass carboy. The air inside the container acts as a stiff spring. Construction C50 = 72.5 liter (19 gallon) vessel C20 = 29 liter (7.5 gallon) vessel C10 = 14.5 liter (3.8 gallon) vessel (Inexpensive 41.6 liter or 11 gallon test lung, measured compliance C28) (Classic adult test lung from from Hill 1965; the 50 liter or 13 gallon glass demijohn inside a wicker basket on the right is the test lung, the device on the left is a pump) A glass container is probably necessary; most plastic or metal containers of this size may not be rigid enough, and as the walls flex, the C value would change in unpredictable ways. (If you really need to use a plastic container — try one half the size first). The test lung may be composed of several smaller containers linked together. If the containers are linked with valves, this allows easily changing compliance by opening or closing valves. In practice, connecting together two or three of the common 5 or 6 gallon carboys such as used in homebrewing or wine making is a good simulation of adult lungs in the C20-C50 range. (Typical brewing carboys) (Vessels of different sizes: 3 to 14 gallons) Optionally the vessel may be filled with copper mesh, about 1kg (2lb) for the largest size or half that for the smaller sizes. The copper mesh is intended to make the compression of air isothermal by adding a large thermal mass that rapidly equilibrates with the air. If the mesh is omitted, compression is adiabatic instead. The mesh mainly has an effect at very low respiratory rates (0–5 breaths/minute); it is generally not necessary if the test lung is used at normal respiratory rates (>5 breaths/minute). Safety Note These suggested vessels are not pressure rated and may fracture at high pressures. The pressure differences used in normal breathing almost never exceed 1 psi, or 70 cmH2O. Compared to bicycle tires or air-powered tools, this is a tiny pressure. There is never a reason to test higher pressures, which may be dangerous. Do not pressurize over 70cmH2O or 1 psi as such pressures may be both unsafe, and are not realistic for lungs. A pressure limiting valve (pop-off valve) set to 1–2 psi is . Use a basket or a cloth wrapping around the vessel to catch any fragments if it breaks, and . required wear eye protection Constructing Compliance with Plastic Sheets and Bags A 2–3 liter inflatable bag (anesthesia bag or similar) sandwiched between two thin flexible sheets of plastic. The sheets of plastic should be connected together on two opposite edges, forming flat springs. The resistance to air flowing in should come (as much as possible) from the bending of the plastic sheets, not from stretching the bag. The bag would typically inflate only to a thickness of 1–3 cm while in use (in other words it is still mostly flat). Adjust the thickness and length of the plastic to get the right C value (shorter lengths = lower C). A reasonable starting point is to use 12x12 inch, 1/8 inch thick plastic sheets made of HDPE, polypropylene or polycarbonate; but the resulting stiffness of the springs is very dependent on the exact thickness and material, so the dimensions would have to be tuned in each case. (IMT SmartLung — commercial version of a bag between flat plastic springs; adjustable length of spring changes C value. Price: $650) To calibrate, iInject of air using a syringe. Measure the pressure rise immediately after. 500ml C50 should be 10 cmH2O C20 should be 25 cmH2O C10 should be 50 cmH2O Extended calibration: inject air in 100ml increments between 0 and 1000ml; measure the pressure rise and draw a curve. An example calibration curve from the two carboy setup above: Constructing Resistance Because resistance is a function of flow, it is measured in pressure divided by flow, or cmH2O/(liter/second). This may be abbreviated with the letter R and a number without stating the units. The typical adult range is around R5 (+-2). Resistance may increase significantly with any conditions that produce impaired airflow, such as asthma. There are two kinds of resistances, linear (pressure proportional to flow rate) and “parabolic” (pressure proportional to square of flow rate). Linear resistances are harder to construct, so the equivalent parabolic resistances are described below. Values used in the ISO standards (for adults) are: R5 — normal R20 R50 — very high In order to construct a flow resistance that models these values, add a thin plate with a small hole to a standard breathing airway. The international standard for airway diameters is 22mm. (A commercial adapter offering three resistances by having thin plates with 3 different hole sizes.) : tube with inner diameter of approx 22mm and at least 40mm long, blocked by a thin plate with a hole that restricts flow. The hole should have the following diameters: Construction R5 = 7.7 mm R20 = 5.6 mm R50 = 3.9 mm In order to calibrate, connect the resistance to a source of pressurized air, with the other side open to the atmosphere. Adjust flow until it is a steady . Hand-held flow meters are relatively expensive but used in many industries, so you may be able to borrow one. If not, you may have to use a large plastic bag of a known capacity and time how long it takes to fill. Measure the pressure at the inlet to the tube (on the side of the plate the air is coming from). 1 liter per second Compared to the atmospheric pressure in the room, the pressures should be: R5 should be 5 cmH2O nominal (2.7 cmH2O for parabolic) R20 should be 20 cmH2O (17.6 cmH2O for parabolic) R50 should be 50 cmH2O (6.8 cmH2O for parabolic, measured at 0.25 liters per second) (note: the specs for parabolic resistance above come from an older standard, ASTM F920–93, table A2.2) In order to perform extended calibration adjust flow to 0.25, 0.5 and 1 liters per second, measure the pressure at each flow and draw a curve. The actual value is likely to be different due to small changes in geometry, etc. Adjust the hole size if needed, or simply use several different sizes with known measured R-values for testing. Other approaches When it comes to simulating lungs, the perhaps obvious DIY approach is to use some kind of expandable or inflatable bag, balloon or similar. This is then loaded in some way e.g. with elastic bands, weights, or by being put under water to make it harder to push air in. The problem with these approaches is that the pressure vs volume curves they have is completely different from the one that the standard test lungs (and real lungs) have. Most elastic bags are not nearly stiff enough; for example anaesthesia bags are much stiffer than balloons, but are still not stiff enough until over-inflated to 2x-5x their rated volume; and additionally, the stiffness changes a great deal once they stretch. Weights, water and similar static loads add a minimum pressure which is required to push any air in; they offset the pressure vs volume curve, rather than changing the slope. Springs, elastic bands and so forth do change the slope, but not in a consistent way — for most of these the slope gets much steeper as the volume gets higher. Trying to measure the R and C values, especially C, is meaningless because the shape of the curve is wrong. This may lead to very misleading conclusions when testing a ventilator prototype. Conclusion It is much more important to have a test lung which behaves in a realistic way, and for which the R and C values are known and within the expected range for real lungs, than it is to have specific R and C values or to use a commercial test lung. Even if a commercial test lung is not available, it is strongly recommended to use the of construction described here — even if the specific sizes or volumes are slightly different — for all but the simplest tests of ventilator function. types Public Invention is a US 501(c)3 public charity. Your supports are attempt to foster tested, reliable, and monitored open-source pandemic ventilators. Public Invention donation References International Standards Organisation. ISO 10651–5–2006 Lung ventilators for medical use — Particular requirements for basic safety and essential performance — Part 5: Gas-powered emergency resuscitators. https://www.iso.org/standard/35975.html Hill DW, Moore V. The action of adiabatic effects on the compliance of an artificial thorax. BJA: British Journal of Anaesthesia. 1965 Jan 1;37(1):19–22. https://academic.oup.com/bja/article/37/1/19/276686 Chatburn RL. Simulation studies for device evaluation. Respiratory care. 2014 Apr 1;59(4):e61–6. https://www.researchgate.net/publication/261515819_Simulation_Studies_for_Device_Evaluation Kenny JE. ICU Physiology in 1000 Words: Shorthand Equations for Respiratory System Power. Jonson B, Svantesson C. Elastic pressure–volume curves: what information do they convey?. Thorax. 1999 Jan 1;54(1):82–7.

How To Make Your Own Test Lungs for Testing Emergency Ventilators

HackerNoon Reporter: Please tell us briefly about your  background. I have built well over 200 websites, having started my first company, a web development agency, at age 16 and eventually growing that to 14 people. I ran that company for 14 years and during that time learned a lot about the craft, but also about workflows and inefficiencies for myself as a designer and developer, but also about the workflows and inefficiencies of the employees, freelancers, and customer teams I worked with. I am very much a tinkerer and optimizer, so finding better ways to do my own work is always on my mind. What's your startup called? And in a sentence or two, what does it do? is the browser for ambitious developers, developers that want to build better websites and also want to be more efficient while doing so. It has tools to help with responsive design, accessibility, and overall site quality. All the tooling doesn’t just tell you where the issues are, it also tells you how to solve them so you won’t have to waste time searching for a fix. Polypane What is the origin story? In 2015 I had just switched over to Sketch and loved the “artboard” feature in it. I could see my designs across different screen sizes side-by-side, easily compare them and make sure everything aligned. It really sped up the design work I was doing. But in the browser, I could only see one size at the same time. That frustrated me so much I decided to hack together a prototype that did what Sketch did: show a site in different screen sizes side by side. I loved using it but didn’t think much of it until I checked my hours and noticed that the same type of projects were suddenly taking just 40% of the time they took me when I didn’t use Polypane. That’s when I knew something was there. So I started sharing it out, getting feedback, and iterating on it. At the end of 2018 I left the web agency I founded and in May 2019 I launched Polypane 1.0. The rest is history. If you weren’t building your startup, what would you be doing? Hopefully working on developer tooling somewhere! I’ve been writing developer tools for as long as I’ve been a developer: CSS frameworks, jQuery plugins, image optimization apps, and more. Even now with Polypane being my full-time work I still tinker away on projects like and , both of which I share out for free. Superposition FixA11y At the moment, how do you measure success? What are your core metrics? It feels like it’s still early days, so active users and MRR. What’s most exciting about your traction to date? Developers are a notoriously difficult market because everyone promises them the world and they’re rightfully skeptical about any claims. They’ll tear right through marketing fluff and that’s great because, being a developer, I don’t like marketing fluff either. So I get to “nerd out” on features and technology choices and share that with my customers. All the growth has been because of customer word of mouth and that has been amazing to see. What technologies are you currently most excited about, and most worried about? And why? Technology has always been a means to an end for me. I don’t enjoy programming at all but I love having built something that didn’t exist when I started and having that do something meaningful. Beyond that though, I think everyone does well to bet on the web (and that includes Electron). What drew you to get published on HackerNoon? What do you like most about our platform? It has a great focus and high-quality articles! What advice would you give to the 21-year-old version of yourself? Go do that startup idea now, go out on your own and believe in yourself. What is something surprising you've learned this year that your contemporaries would benefit from knowing? The more polished you make something, the harder it is for people to see the quality that went into it, but also the more they’ll appreciate it when you tell them about it. Don’t be afraid to tell them about that. Polypane was nominated as in Startup of the Year hosted by HackerNoon. one of the best startups in Zoetermeer

Polypane's Founder Kilian Valkhof on Building the Browser He Wished He Had

Machine Learning for Business Bolstering an organization’s productivity is probably the first application of Machine Learning (ML) in business that comes to mind. Any company that switches to a more data-driven approach, would first and foremost utilize Machine Learning to help themselves produce more or, more accurately, produce efficiently. Evaluating and quantifying the end result is pretty straightforward. However, there has been an increasing interest in the applications of ML in the human resources department. Gathering employee (or potential employee) data can be a really powerful tool in the hands of HR practitioners to help them accomplish their tasks quicker and in some cases (like hiring) in a less biased manner. Applications of ML in HR Integrating ML  in HR processes is not a distant future theory. It has been embraced by many enterprises already, and the trends show that it will soon become the norm. Most of the use cases fall into three major groups: Hiring Hiring the best available talent for a job is a two-way process. Certain Natural Language Processing tools can be a useful aid to compose a job description that will sound really attractive and interesting to the candidate. On the other hand, ML can make the HR practitioner’s life easier when going through hundreds (or even thousands) of resumes, and filter out those that are not a great match for the role. Having an algorithm to do this can greatly reduce the unavoidable human bias as well as the amount of time required. An interesting case study is Unilever. They switched to a more data-driven hiring model, with the help of HireVue and this led to £1M+ annual cost savings, 96% candidate completion rates, 90% reduction in time to hire, and  16% increase in new-hire diversity. You can read the case study . here Engagement An ML-based system can process data collected via various intra-company sources and provide insights on how engaged the employees are in their work. These findings are easily missed by the human eye, especially for very big companies. Personal Development and Learning There are studies available that demonstrate how important development opportunities are for employee happiness. In some age groups (such as the millennials), a robust personalized framework for growth is shown to be more important than the salary. The same applies to the employer. It is for their benefit to keep their staff up-to-date with the latest trends and technologies. Blindly spending money on training their employees is inefficient. Two steps are necessary here: Identify the skill gap on an individual level and a) Create a learning path for them. b) “Individual” is a key word. The development plan must be as personalized as possible. There is no one size fits all solution. This is where ML can help, by analyzing the employee’s data and highlighting the missing or weak skills followed by generating a plan of action and monitoring the progress made. As an example, you can read the of Amadeus, a travel technology company, that designed and built the  Valamis-based Amadeus Learning Universe (ALU) global learning platform in order to train Amadeus employees, partners, and clients worldwide. case study Attrition You invested a lot of time and money in finding and attracting a great candidate and they are performing as expected. You want to make sure that they are happy and committed to the company. Losing great talent can not only be frustrating but they can also be really expensive to replace, especially for senior or managerial positions. There is a series of costs involved in that, ranging from finding new people to onboarding and training them. ML algorithms can help with providing a better vision of which are the most important factors that led to employees leaving in the past, as well as calculating the likelihood of a specific individual leaving in the future, based on past data. In this story, I will focus on job attrition and will demonstrate how we can use ML to help analyze the data and make predictions. The PyCaret library will be used to speed up the process. PyCaret defined the citizen data scientist as a person: Gartner “Who creates or generates models that use advanced diagnostic analytics or predictive and prescriptive capabilities, but whose primary job function is outside the field of statistics and analytics.” PyCaret is a library that is inspired by the emerging role of citizen data scientists. It is the Python version of the Caret package for R, essentially a wrapper around some of the most popular machine learning libraries, such as scikit-learn, XGBoost, and many others. It greatly simplifies the lifecycle of an ML experiment from model training to deployment. Behind the scenes, PyCaret takes care of a great amount of the necessary coding, exposing easy-to-use functions to the user. This leads to a great reduction of the lines of code that actually need to be written. From the documentation page: PyCaret is an open-source, low-code machine learning library in Python that automates machine learning workflows. It is an end-to-end machine learning and model management tool that speeds up the experiment cycle exponentially and makes you more productive. The code Setup and objective I will be using the IBM HR Analytics Employee Attrition & Performance available from Kaggle, which contains a mixture of numeric and categorical variables. The objective is to predict if an employee will leave the company (attrition). I will use Google Colab to train the model and perform the analysis. For this specific example, the dataset is assumed to be saved at this location: dataset This is a typical binary classification task, where the target variable (Attrition) takes 0/1 values. /content/drive/MyDrive/Data/HR-IBM_ATTRITION/WA_Fn-UseC_-HR-Employee-Attrition.csv Install / import packages and read the data !pip install pycaret shap

from pycaret.classification import *
import pandas as pd
from pycaret.utils import enable_colab
enable_colab()

df_ibm = pd.read_csv('/content/drive/MyDrive/Data/HR-IBM_ATTRITION/WA_Fn-UseC_-HR-Employee-Attrition.csv')
df_ibm.info() Exploratory data analysis findings After performing EDA, the major insights were: There are no missing values Some of the features have a constant value. These can be dropped (‘Over18', 'EmployeeCount', 'StandardHours', 'EmployeeNumber') There are a lot of numeric features that do not follow a normal distribution Some numeric features have only a few unique values There is some multicollinearity present for some of the features The   variable to be predicted is highly imbalanced So we will start by dropping the columns that don’t offer anything to the model df_ibm.drop(['Over18','EmployeeCount','StandardHours','EmployeeNumber'],axis=1,inplace=True) Setup PyCaret environment Before running any model training, the environment must be set up. This is super easy, all we need is to call the function, but first, we need to take care of how the variables will be interpreted. PyCaret will infer the various variable types (numeric, categorical, etc). We can override this behavior by providing our own association of variables with data types. Let’s create these lists and dictionaries first and have them ready to be passed as arguments to the setup function. setup() cat_vars = ['Department','EducationField','Gender','JobRole','MaritalStatus','OverTime',]
ordinal_vars = {
    'BusinessTravel' : ['Non-Travel','Travel_Rarely','Travel_Frequently'],
    'Education' : [str(x) for x in sorted(df_ibm['Education'].unique())],
    'EnvironmentSatisfaction' : [str(x) for x in sorted(df_ibm['EnvironmentSatisfaction'].unique())],
    'JobInvolvement' : [str(x) for x in sorted(df_ibm['JobInvolvement'].unique())],
    'JobLevel' : [str(x) for x in sorted(df_ibm['JobLevel'].unique())],
    'JobSatisfaction' : [str(x) for x in sorted(df_ibm['JobSatisfaction'].unique())],
    'PerformanceRating' : [str(x) for x in sorted(df_ibm['PerformanceRating'].unique())],
    'RelationshipSatisfaction' : [str(x) for x in sorted(df_ibm['RelationshipSatisfaction'].unique())],
    'StockOptionLevel' : [str(x) for x in sorted(df_ibm['StockOptionLevel'].unique())],
    'TrainingTimesLastYear' : [str(x) for x in sorted(df_ibm['TrainingTimesLastYear'].unique())],
    'WorkLifeBalance' : [str(x) for x in sorted(df_ibm['WorkLifeBalance'].unique())]

}
numeric_features = ['DailyRate','DistanceFromHome','Age','HourlyRate','MonthlyIncome','MonthlyRate','NumCompaniesWorked','PercentSalaryHike','TotalWorkingYears','YearsAtCompany','YearsInCurrentRole','YearsSinceLastPromotion','YearsWithCurrManager'] We are now ready to set up the environment. experiment = setup(df_ibm,target=‘Attrition',
                   categorical_features=cat_vars,
                   train_size=0.8,
                   ordinal_features=ordinal_vars,
                   remove_multicollinearity=True,
                   multicollinearity_threshold = 0.9,
                   transformation=True,
                   normalize = True,
                   numeric_features=numeric_features,
                   session_id = 42) Let me explain some of the parameters of this method: setup() : Name of the target column to be passed in as a string. The target variable can be either binary or multiclass. target : Proportion of the dataset to be used for training and validation. By default, PyCaret will split the dataset to train/test splits when running setup on a 70/30 ratio. We have overridden this value. train_size : If the inferred data types are not correct or the silent param is set to True, categorical_features param can be used to overwrite or define the data types. It takes a list of strings with column names that are categorical categorical_features : Encode categorical features as ordinal. For example, a categorical feature with ‘low’, ‘medium’, ‘high’ values where low < medium < high can be passed as ordinal_features = { ‘column_name’ : [‘low’, ‘medium’, ‘high’] }. : If the inferred data types are not correct or the silent param is set to True, numeric_features param can be used to overwrite or define the data types. It takes a list of strings with column names that are numeric. ordinal_features numeric_features : When set to True, it transforms the numeric features by scaling them to a given range. The type of scaling is defined by the normalize_method parameter (default is score). normalize : When set to True, it applies the power transform to make data more Gaussian-like. The type of transformation is defined by the transformation_method parameter (default is yeo-johnson). transformation : Controls the randomness of the experiment. It is equivalent to ‘random_state’ in scikit-learn. When None, a pseudo-random number is generated. This can be used for later reproducibility of the entire experiment. session_id : When set to True, features with the inter-correlations higher than the defined threshold are removed. When two features are highly correlated with each other, the feature that is less correlated with the target variable is removed. Only considers numeric features. remove_multicollinearity : Threshold for correlated features. Ignored when remove_multicollinearity is not True. multicollinearity_threshold Model creation and training To get a list of all the available classification models, we can simply run : which returns: models() We can conveniently see how all the above models perform on the train dataset: top3 = compare_models(n_select=3,sort='AUC') This function trains and evaluates the performance of all estimators available in the model library using cross-validation (10 fold is the default). The output of this function is a score grid with average cross-validated scores. In this example, I chose AUC to be the metric that models are sorted by. The variable is now a list that holds the 3 best performing models, and they can be referenced like normal Python list elements. Can you imagine? With a single line of code, we trained all the models and generated a metrics grid. Nice. top3 It is worth mentioning here that PyCaret allows individual model creation very easily, as well as various ensembling techniques (boosting, bagging, blending, stacking). Check the documentation on how to perform these steps. Just to give you an idea, we could blend the top3 models like this : blend = blend_models(top3) But for now, we will use the top model, Logistic Regression. The next step would be to try and fine-tune the model. PyCaret has us covered. tuned_lr = tune_model(top3[0],n_iter=70,optimize=‘AUC') This function tunes the hyperparameters of a given estimator. The output of this function is a score grid with CV scores by fold of the best selected model based on optimize parameter The parameter is the number of iterations in the grid search. Increasing n_iter may improve model performance but also increases the training time. n_iter By default, tune_model performs scikit-learn . random grid search Model performance We are now ready to evaluate the model’s performance on the holdout set (remember, PyCaret created one when we ran the setup).
This is again very easy: pred_holdout = predict_model(tuned_lr) PyCaret takes the burden off us to create the plots that would visually depict the model’s performance. Let’s see some examples: plot_model(tuned_lr,’confusion_matrix') plot_model(tuned_lr,'auc',save=True) plot_model(tuned_lr,'learning',save=True) Before saving the model, one last useful step is to run the function. Once the predictions are generated on the hold-out set using and we have chosen to deploy the specific model, we want to train our model for one final time on the entire dataset including hold-out. finalize() predict_model final_tuned_lr = finalize_model(tuned_lr) To save the final model, we need to run: save_model(final_tuned_lr, 'final_tuned_lr') Which saves the entire transformation pipeline and trained model object as a transferable binary pickle file for later use. As you can imagine, loading the saved model is as easy as: final_tuned_lr_loaded = load_model('final_tuned_lr') Model interpretation - SHAP values There is another feature I would like to show you, which is model interpretation. The implementation of the interpretations is based on (SHapley Additive exPlanations).  SHAP values are a widely used approach from cooperative game theory that come with desirable properties. They help us demystify how the model works and how each feature affects the final prediction. However, only tree-based models for binary classification are supported  (Extra Trees Classifier, Decision Tree Classifier, Random Forest Classifier, and Light Gradient Boosting Machine). Let’s quickly demonstrate how straightforward it is for PyCaret to interpret a model. We will train and tune a Random Forest Classifier first. Instead of using the compare_models() function, we can create individual models like so: SHAP rf = create_model(‘rf') and then tune it: tuned_rf = tune_model(rf,optimize=‘AUC') Next, we will generate the SHAP values for each feature: interpret_model(tuned_rf,save=True) Now, how do we read this plot? The logic is that: For each feature observation (dots), the red color represents high values, while the blue color represents low values. If the dots are more spread to the right, this means that the feature has a high effect on the positive class (Attrition = 1). Similarly, if the dots are spread to the left, the feature has a negative effect on attrition. Looking at the graph, we can conclude that: Features that have a great effect on attrition Low monthly income Low total working years Low age High overtime hours Features that have a medium effect on attrition Low years at the company A high number of companies worked before High distance from home High years since last promotion Further reading I encourage you to check the PyCaret documentation page and try some of their tutorials. I just covered a small fraction of what the library can do here. PyCaret Links https://greenice.net/how-leading-companies-use-machine-learning-to-solve-major-hr-challenges-greenice/ https://towardsdatascience.com/machine-learning-in-human-resources-applications-and-trends-7f7a1473b222

POLYCARE

polycare.de

11-50 employees

Since 2010

COMPANY RANKING

POLYCARE

EVERGREEN INDEX #8579

Representing this company? Click here to claim and customize this page!

Polycare WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!

POLYCARE

polycare.de

11-50 employees

Since 2010

COMPANY RANKING

POLYCARE

EVERGREEN INDEX #8579

RELATED COMPANIESRANK

STORIES ABOUT POLYCARE

Representing this company? Click here to claim and customize this page!

Polycare WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!