Temporary Tattoos Perceived by the Skin-IEEE Spectrum

2021-11-24 01:57:27 By : Mr. Henry Chen

The IEEE website places cookies on your device in order to provide you with the best user experience. By using our website, you agree to the placement of these cookies. To learn more, please read our privacy policy.

I turned the key to start the small Ford SUV that I rented to visit the University of Illinois at Urbana-Champaign, and a message flashed on the dashboard: "Low tire pressure." I ignored it. I’ve been in my own car for 12 years; I’m not used to cars that monitor my health. However, it turns out that Ford is not joking. The next morning, I found that the tire of the car was flat.

Ironically, it was this idea that brought me to Illinois. I came here to see the bold vision of John Rogers, professor of materials science at the University of Illinois: He believes that one day, our bodies will have sensors that can send information to mobile phones, just like the way cars send. . The sensor powers the vehicle's computer.

Although all bio-seals have some common features—they stretch like skin, include flexible circuits, and can be wirelessly powered—but different functions require different sensors. The butterfly sensor on the left is used to monitor the sun's ultraviolet (UV) exposure, the sensor in the middle uses sensitive dyes to detect chemicals in sweat, and the sensor on the right uses an electronic circuit to measure blood pressure.

We have already taken the first step in this direction. Many of us now wear fitness bands that can track our activity and heart rate. If we don’t move for a while, we think we’re sleeping. But most of these bands are not completely fashionable or unobtrusive, so even the more stubborn ones among us sometimes take them off. The information they provide is interesting, but almost unimportant: they cannot detect signs of your illness, nor can they tell your doctor anything they need to know, let alone replace an office visit.

But they have no reason not to, Rogers said. Think about your last medical examination. Your doctor checked your pulse, temperature, blood pressure, and perhaps blood oxygen. If there are any abnormalities, you may be sent for further tests-it may be your heart electrocardiogram, blood tests to check for diabetes, if you have an electromyogram of muscle weakness, or maybe even polyconductivity in a sleep laboratory Sleep chart to check for apnea. All of these tests require professional and expensive equipment, well-trained medical technicians or invasive probing.

These tests-and more-can be done with lightweight, durable and comfortable sensors that you can wear on your body for weeks at a time. Not an unattainable dream: As of press time, many sensors developed by Rogers research team members have entered or are about to enter clinical trials in the United States and Europe, and the first commercial versions are expected to be available by the end of this year.

Rogers said that these sensors are so skin-like that you won't notice that you are wearing them-I don't have to believe him. I have worn it on the inner forearm for more than a week. This version is a test unit. When triggered by an Android smartphone, it will only send a greeting; a device with a biosensor has not been provided to reporters.

Although simple, my sensor makes me very happy. When I was living my life-bathing, sleeping, and exercising, it stuck to my arm unobtrusively and tenaciously. It also made me think about how future versions of these sensors will improve our lives—not in ten years, but in a few years, Rogers promised.

The Illinois team is not the only one trying to make skin-like electronic products. Takao Someya is leading a team at the University of Tokyo dedicated to the development of electronic skins made of organic semiconductors and carbon nanotubes. Zhenan Bao of Stanford University is also working with organic semiconductors to develop an electronic film that is as sensitive as human skin and can be applied to robot limbs. Researchers at the University of California, San Diego are developing inks that allow scientists to map sensors directly to the skin.

But Rogers' skin-like sensor is expected to be the first sensor to walk out of the laboratory and enter our body. In 2008, Rogers and Roozbeh Ghaffari co-founded a company called MC10 in Cambridge, Massachusetts, which transformed his team's research on stretchable electronics into commercial health products. Today, MC10 has approximately 60 full-time employees, US$60 million in venture capital and corporate investment, and a product on the market: Checklight, a capless device used to accurately measure acceleration when an athlete’s head impacts. It is not a skin-like sensor patch, but it can be bent to conform to the shape of the body part. (Rogers serves on the board of MC10 and helps plan the company's research and technical work with Ghaffari. Ghaffari is now the chief technology officer of MC10.)

MC10 started making the first skin patches at the end of 2012-the company calls them Biostamps. Most of these early products are used for internal development or joint development work with partners. MC10 began developing a new generation of technology at the end of 2014; most of these bio-seals will now be handed over to medical researchers for clinical trials. Consumer Health Biostamps is also being developed for companies targeting their own special fields. For example, a cosmetics company may package the Sunshine Monitoring Biostamp with sunscreen, or a pharmaceutical company may package the Sports and Temperature Monitoring Biostamp with the drug package.

The basic Biostamp is a thin sticker, about the size of 10 pence in the United Kingdom or 25 cents in the United States. It looks like a temporary tattoo that a child might get at a birthday party, but because its design is mechanically similar to the skin, once it is applied, the wearer cannot really feel it. Biostamp can contain hundreds of thousands of transistors, as well as resistors, LEDs, and RF antennas. It is waterproof and breathable, and mass production only costs tens of cents. It can be worn for about a week, and then the normal shedding of skin cells begins to force the thin matrix to peel off the skin, just like an early sunburn.

Biostamp consists of a stretchable circuit and is supported by a very thin rubber sheet. To make these circuits, Rogers and his colleagues in Illinois first made transistors, diodes, capacitors, and other electronic devices on wafers of any common semiconductor material. They usually use silicon, but gallium arsenide or gallium nitride can also be used. These are not ordinary semiconductor wafers; they are a bit like Oreo cookies for semiconductor wafers. They have a thin top layer of semiconductor material, a thicker bottom layer of the same material, which acts as a rigid support during the manufacturing process, and a sacrificial layer of different materials in between. In the case of silicon wafers, the sacrificial layer is silicon dioxide. After the equipment is manufactured, the chemical bath will erode the center layer and release the thin top layer.

Then a stamp made of soft silicone is pressed onto the wafer. The raised area on the stamp lifts the selected electronic device, like a rubber stamp sucking ink from the printing pad. After picking up the equipment, the silicone seal deposits them on a temporary substrate, usually a plastic-coated glass plate. The board then connects the device with a copper conductor in the form of a serpentine coil through a standard photolithography process, making the connection stretchable.

The next step is to transfer the interconnection device from the plastic coated glass to the consumer-a thin rubber sheet that has been attached to the plastic backplane with a layer of adhesive in between. To this end, a machine pushes the rubber toward the device and coil array that still clings to the plastic-coated glass. The final chemical bath dissolves the plastic between the electronic circuit and the glass, allowing the circuit to adhere to the rubber. The final step occurs when Biostamp enters the user's hand-the user exposes the adhesive and sticks the rubber-backed electronic device to the skin.

In many Biostamps, all electronic products are created using this process. However, in some cases, the Biostamp design contains an unpackaged microprocessor, which the researchers thinned to 5 to 10 microns. The circuit is covered with a few micrometers of flexible resin for waterproofing. However, for now, most Biostamps are not yet equipped with mature microprocessors. Most people who are being tested now just collect and transmit data; the analysis takes place elsewhere, usually on a smartphone or tablet.

Biostamp powers itself by harvesting energy from Near Field Communication (NFC) radio signals (usually from the wearer's mobile phone). It communicates with the phone in the same way. NFC sends data at 13.56 MHz and is a feature of almost all current smartphones that use it for wireless payment solutions. Currently, stamps are only available for Android phones, but the hardware is compatible with the type of NFC technology on newer iPhones.

Biostamp Brigade: These stamps include a variety of ultra-thin unpackaged electronic products, flexible circuits and sensitive dyes. They collect energy and communicate wirelessly, and can support various sensors to allow monitoring of different bodily functions. Photo: Randy Klett

The stamp uses an induction coil to convert the radio frequency energy received by the antenna into electrical energy. When it is about one meter away from a mobile phone that transmits NFC signals, Biostamp can generate tens of milliwatts of power. For longer-distance power collection, Biostamp can be built to receive radio signals with frequencies between 1 and 2.5 GHz from a transmitter a few meters away.

Although Rogers' team and MC10 have built and tested retractable batteries and supercapacitors, the current tattooed version of the technology does not store energy. But in a ward, such as an NFC transmitter under the bed or a farther RF transmitter in the corner, Biostamps can operate continuously and indefinitely.

Currently, Rogers and his students are evaluating stretchable sensors that can measure body temperature, monitor UV exposure, and check pulse and blood oxygen levels. They are also developing sensors that can track changes in blood pressure, analyze sweat, and obtain signals from the brain or heart for EEG and ECG. Rogers said that all of these sensors are designed to measure with sufficient accuracy to be used in a medical environment-much higher than the standards required to build a typical consumer wearable device.

Achieving this standard in thin and stretchable devices forced Rogers and his team to rethink how to perform medical measurements. Consider blood pressure: blood pressure is usually measured by putting a cuff on the patient's arm and inflating the cuff until it cuts off the blood flow. Then the cuff is gradually deflated until blood begins to flow through the blood vessels in the arm again; this gives a systolic pressure reading, which is a measure of pressure during systole. The cuff continues to deflate until the doctor can no longer detect the sound of blood flowing. This produces a diastolic blood pressure reading, which is the pressure on the heart muscle as it relaxes between heartbeats.

A tiny Biostamp cannot cut off the blood flow. However, it can measure the pulse between two points at a distance of only about one centimeter. With this information, the smartphone can calculate a physiological index called pulse wave velocity, which changes with blood pressure. Therefore, researchers in Rogers group Tony Banks, SeungMin Lee and Matt Pharr are developing two types of Biostamp pulse detectors. A kind of light used; it alternately flashes red and infrared LEDs, and uses a photodetector to pick up the light reflected by the skin under the Biostamp. Because deoxygenated blood absorbs more red light and oxygenated blood absorbs more infrared light, fluctuations in these levels produce a waveform that represents the heartbeat. This is basically how the latest fitness band detects pulses, although the Biostamp version can get a more stable signal because the skin will not move under it. Another type of pulse detector being developed uses piezoelectric strain sensors to monitor the stretch and relaxation of the patch as it responds to the blood flowing through the blood vessels below it. In this scenario, greater stretching translates to higher pressure.

In either case, these pulse measurements converted to pulse wave velocity can only tell the wearer how blood pressure changes-they do not give a baseline blood pressure reading. But for patients who need to monitor blood pressure closely, simply tracking changes is essential. Patients do not need to see a nurse for blood pressure checks or use bulky home monitoring equipment every day. Instead, they put on a new Biostamp every week or every two weeks, calibrate it, and then scan the seal with their mobile phone to get possible readings. Automatically sent to the doctor.

Another promising version of Biostamp will monitor sweat. Daeshik Kang and Aheyon Koh, postdoctoral fellows in Rogers' team working with researchers Banks, constructed a Biostamp with microfluidic channels that can absorb sweat along a calibrated path. The chemically sensitive dyes on the path change color with sweat, while other dyes on the patch change color in response to glucose, lactic acid, chloride, and sodium. When scanned with a smartphone, the circuit on the patch activates a mobile phone app that analyzes color changes and provides suggestions, such as "time to replenish water," or, for women who are monitoring for gestational diabetes, "it's time I went to see a doctor." "MC10 researchers, in collaboration with Rogers' team, are also investigating possible ways to use this sweat-derived data to monitor heart health.

Unlike all-electronic wearable sweat monitors developed elsewhere, these chemically sensitive Biostamps cannot be reused. At the end of your workout, competition, or stress test, when you stop sweating, you will not be able to reset the patch, so you must throw it away. However, they will be very cheap and you won't care.

Rogers believes that Biostamps will become the main product of the hospital. This year, a preliminary trial began in the neonatal intensive care unit of the Carr Foundation Hospital in Urbana, Illinois, where doctors are using Biostamps to monitor newborns’ temperature and other vital signs. The doctor attached Biostamps to each baby's arms, legs, forehead and chest, while the NFC antenna under each incubator provided power to the device.

Following the news of his Biostamps, Rogers and MC10 received not only requests from doctors and trainers, but also requests from government officials and business executives. Many people discovered his work by reading one of more than one hundred papers that appeared in scientific journals in recent years.

Thin and flexible: This Biostamp is a test bed for various stretchable devices; it includes an array of transistors, diodes, capacitors, inductors, LC oscillators, temperature sensors, strain gauges, LEDs, induction coils and simple antennas. Photo: Randy Klett

"I read his paper in Science four years ago," said Guive Balooch, global vice president of new technologies at the hair care and cosmetics giant L'Oréal. "We went to him because measuring the skin and understanding the changes over time can help us identify and test the product."

L'Oréal is now working with the Biostamp team to develop a skin moisture sensor that monitors how heat passes through the skin under the patch. The device on the patch generates a small amount of heat, which is detected by the temperature sensor on the same patch. L'Oréal hopes to eventually use this information to test the efficacy of its products; the patch can track changes in hydration as people use its products over time, as well as more general changes that occur as the skin ages. The company's researchers have conducted an experiment on 20 subjects, each of whom wore 6 Biostamps. The initial research only established the correlation between hydration, temperature, skin thickness, and heat transmission through the skin. Balooch predicts that in 5 to 10 years, the technology will play a bigger role. "I would love to see the beauty stickers on someone to give them skin care advice," he said. L'Oréal is also funding research by Biostamps, which measures UV exposure and sends a signal when it needs to reapply sunscreen.

In the more than a week I wore the lean Biostamp, I found myself wanting to roll up my sleeves and show it off. After seeing numerous demonstrations of potential applications in Rogers Lab, and especially after recalling L’Oréal’s research, I was even annoyed that I had to do things the usual way. One hot afternoon, I was sitting in the sun, arguing whether to reapply sunscreen, looking at the patch on my wrist, I thought, "You can tell me this, you know."

Later in the same week, I felt as if I had a cold, and I went to look for a thermometer. I stared at my Biostamp again and hoped it was the temperature sensor I saw in Illinois. I have worn the Fitbit Flex for more than a year. Ever thought it was very fashionable? It looks big and unwieldy to me now.

Other researchers are studying the possibility of using temperature-sensing Biostamps to measure the mental stress of air traffic controllers (the more difficult the mental task, the lower the temperature of the hands) and the possibility of using fever Biostamps to push drugs into the skin. In a clinical trial at Northwestern University School of Medicine in Chicago, the research team tested Biostamps, which can measure temperature and heat flow in tissues to monitor wound healing. The bio-seal, which is attached behind the ear to measure the electrical activity of the brain for sleep research, will undergo clinical trials at the Carl Foundation Hospital this spring. This method is much simpler than wired sensors that are commonly used today.

Of course, any device that collects health data must ensure patient privacy, so any application that uses Biostamp data must comply with the security requirements of the Health Insurance Portability and Accountability Act of the United States and similar privacy regulations around the world. But unlike other data collection devices, Biostamps also has the potential to make health information more secure. Because it cannot be removed without being destroyed, Biostamp can be a physical key used to control access to data, whether in the patient's smartphone or the nurse's workstation.

Early tests of different Biostamps showed encouraging results, but they were labor-intensive and required Rogers' team to design each sensor from scratch. Therefore, although medical researchers have many ideas on how the proposed sensor can help patients, it may take months or even years to create it. The stamp’s limited memory and external power supply limit the types of applications that are feasible.

To solve this problem, Rogers and MC10 developed a version of Biostamp, which is larger (about the size and shape of a band-aid, although slightly thicker), reusable, and equipped with various sensors, batteries, and memory. The patch can be placed on various parts of the body, and the signals it collects can be analyzed by a smartphone or tablet application. This reusable version includes off-the-shelf chips for NFC or Bluetooth low energy communication (chips peeled from the package) and multiple sensors, small square lithium-ion batteries, and a device for connecting these components together The snake coil Rogers invented the stretchable circuit used in the tattoo-like Biostamp.

Researchers have begun to use these patches in clinical trials and use them as a platform to develop new applications, some of which may migrate to the smaller skin-like Biostamp. MC10 plans to sell these larger, reusable Biostamps in 2016 as a competitor to today's health tracking devices. As long as its battery can be charged, this gadget will work. Depending on how often you end up charging, this may take two years or more.

With the imminent commercialization of skin-like wearable electronic products, Rogers is shifting his attention from electronic products that can be sensed from the outside of the body to electronic products that can be worn inside. He is collaborating with researchers at the University of Pennsylvania to study an array of approximately 400 electrodes that can be covered on brain tissue to map the activity that signals epileptic seizures. Researchers have already tested cats and will soon start primate experiments.

Other researchers are using hearts removed from organ donors to test biological seals that can be laminated directly to the surface of the heart. This type of sensor — and the mesh sensor that ultimately completely envelops the heart and extracts energy from its beating — will provide detailed information about arrhythmia, and can provide finer details for current pacemakers that monitor individual points in organs. control.

Recently, Rogers began to consider new challenges. Certain parts of the body, such as the brain and heart, have twists and cracks, and need a 3-D solution instead of his 2-D solution. "We not only hope to transform the circuit from a flat wafer into a thin and soft sheet that can be wrapped on a complex surface," he said, "but also to promote their self-assembly into an open 3-D format, which has a fully permeable biological system. Arrays of filaments and interconnected structures. This capability will allow us to enter a whole new field of biological integration."

Such applications are still more than ten years away. But Rogers' near-term vision is also very compelling. The passionate inventor fully predicts that within ten years, almost everyone in the developed world will wear one or more Biostamps, at least at some point.

Fast forward to 2025. At this time, if the dreams of Rogers and MC10 come true, babies in developed countries will be affixed with a few Biostamps labels at birth. One worn on the wrist or ankle will be used as a high-tech hospital bracelet-and it is more difficult to lose than today's plastic bracelets. Others on the torso or arms will allow the nurse to quickly scan temperature, oxygen saturation and pulse without disturbing the sleeping baby. The baby's mother will also wear some Biostamps so that the nurse can monitor her vital signs when she recovers; the self-inflating blood pressure cuff will no longer wake up an exhausted new mother.

In the same hospital, patients with heart disease wear the vital signs Biostamps and the other two on their ankles to check for swelling, which is an early sign of heart failure. After the heart disease patient is discharged from the hospital, the stamp will continue to be monitored at home. Even nurses will wear Biostamps to enable them to open doors and log in to their computers; these devices will provide greater security than scanning cards or passwords.

Outside, joggers who run through the hospital will wear Biostamps to monitor their progress towards their fitness goals. Although a typical jogger may not suffer from a serious illness, Fitness Biostamps will detect early signs of heart problems or movement disorders such as Parkinson's disease and advise the wearer to see a doctor.

At the same time, passengers lining up to board the cruise ship at a nearby port will be busy adding another Biostamp to the clothes they usually wear. There is a cruise ship logo on it, which will serve as a secure ID to allow them to board the ship. It can also open their hatch so that they can recharge their drinks at the bar, even monitor their sun exposure during the week, and warn them when they need to reapply sunscreen.

Considering that most people have not even seen Biostamp, this kind of Biostamp that is about to become popular may be hard to imagine. But technology sometimes surprises us. Ten years ago, iPhone and Android were still in the design stage, and now we rely on them to continuously obtain information about the outside world. Ten years later, if Rogers succeeds in creating the world of Biostamp, we will also understand our internal world.

This article was originally published under the name "Give your body a'check engine' light".

Tekla S. Perry is a senior editor of IEEE Spectrum. For more than 30 years, she has been working in Palo Alto, California, reporting on the people, companies, and technologies that make Silicon Valley a special place. As a member of IEEE, she holds a bachelor's degree in journalism from Michigan State University.

It provides courses on entrepreneurship, finance, etc.

Professionals who specialize in engineering and technical management must understand interdisciplinary concepts and contribute to multi-functional teams. Although technical expertise is important, it is not enough for long-term career development and success.

Many engineers and technical professionals lack important skills. With the recent shift to remote work, many organizations have not done enough to train them. According to LinkedIn Learning, the consequences are obvious. LinkedIn’s guide "How to Attract and Retain Top Talent in Learning Plans" states that employees who think their career goals have been put on hold are 12 times more likely to consider leaving.

By investing in employee leadership development programs, organizations can retain the best talent. In a CareerBuilder survey of employers on the impact of hiring people with advanced degrees, 32% thought the retention rate had increased. This is why IEEE cooperates with Rutgers School of Business to provide the only mini MBA course designed for engineers and technical professional teams.

"This course has a reasonable structure, allowing us to appreciate the world outside the engineering field."

Recently named as one of the top three mini-MBA courses by IEEE and Forbes | Rutgers Online Mini-MBA for Engineers is a completely virtual course that provides basic courses traditionally taught in the MBA program. Courses cover accounting, business communication, business ethics, entrepreneurship, finance, managerial economics, management, marketing, operations and strategic management.

Completing the plan allows learners to:

The following are the comments of some project graduates:

"The reason I took this course is to better understand the'other side'," said Sohaib Qamar Sheikh, a technical assistant at a large commercial real estate development and investment company in the UK and an IEEE senior member. "This course is well-structured and gives us a taste of the world outside of the engineering field. It helps me better understand other aspects related to our business products."

"I decided to participate in the program for the following reasons: it is cost-effective, rich in content, and flexible to adapt to my schedule," said IEEE member Anis Ben Arfi, a system engineer at ADI. "Through my Mini MBA course, I learned a variety of skills and improved my potential to easily handle business operations. I have a better understanding of trade secrets and patents, more familiar with product life cycles, different metrics for evaluating and measuring customer experience, and agility Project management process. My advice to aspiring applicants is: if you have the opportunity to participate in this journey, please seize it."

Registration is now open to individuals interested in attending next year’s conference. There are two sessions available. One starts in March; the other starts in September. The registration deadline for the March conference is February 4, and the registration deadline for the September conference is August 15. Individuals interested in registering can contact IEEE account experts.

IEEE | ​​Rutgers Online Mini-MBA for Engineers is also available to organizations interested in gaining access to groups of 10 or more people. If you are interested in group access and pricing, including custom apex options designed for your organization's needs, please contact an IEEE account specialist.

Interested in learning more about leadership? Here are two free IEEE on-demand virtual events that can help future leaders bridge the gap between business and engineering as they prepare to grow into a managerial role:

Nvidia’s Earth-2 will use machine learning to improve model resolution, while the EU’s efforts take a different approach

Edd Gent is a freelance technology writer based in Bangalore, India. His work focuses on emerging technologies across computing, engineering, energy, and biological sciences. He is @EddytheGent on Twitter and emailed at edd dot gent at Outlook dot com. His PGP fingerprint is ABB8 6BB3 3E69 C4A7 EC91 611B 5C12 193D 5DFC C01B. His public key is here. DM signal information.

A powerful climate model helps eliminate any uncertainty about the scale of the climate crisis facing the world. But these models are large-scale global simulations and cannot tell us how climate change will affect our daily lives or how to respond at the local level. This is where the digital twin of the earth can help.

Digital twins are virtual models of real-world objects, machines, or systems that can be used to evaluate the performance of real-world objects, diagnose or predict failures, or simulate how future changes will change their behavior. Typically, digital twins involve digital simulation and real-time sensor data from real-world systems to keep the model up to date.

So far, digital twins have been mainly used in industrial environments. For example, a digital twin can monitor the power grid or manufacturing equipment. However, there is increasing interest in applying similar ideas to the field of climate simulation to provide a more interactive and detailed way to track and predict changes in the systems that drive the Earth’s climate (such as the atmosphere and oceans).

Now, the chip manufacturer NVIDIA is committed to building the world's most powerful supercomputer, specifically for simulating climate change. At the company's GPU technology conference, CEO Jensen Huang said that Earth-2 will be used to create a digital twin of the earth in Omniverse-this is a virtual collaboration platform and an attempt by NVIDIA in the meta universe.

"We may finally have a way to simulate the Earth's climate in 10, 20 or 30 years, predict the regional impact of climate change, and take action to mitigate and adapt before it is too late," Huang said.

The announcement did not disclose details. An Nvidia spokesperson said that the company is currently unable to confirm what the computer's architecture is or who can access it. But in his speech, Huang emphasized the company's belief that machine learning plays an important role in improving the resolution and speed of climate models and creating a digital twin of the earth.

Today, most climate simulations are driven by complex equations that describe the physics behind key processes. Many equations are computationally expensive to solve, so even on the most powerful supercomputers, models usually only reach a resolution of 10 to 100 kilometers.

Huang said that some important processes, such as the behavior of clouds that reflect solar radiation back into space, only operate on a scale of a few meters. He believes that machine learning can help here. While announcing Earth-2, the company also launched a new machine learning framework called Modulus, designed to help researchers train neural networks to simulate complex physical systems by learning from observation data or the output of physical models.

"The resulting model simulates physics 1,000 to 100,000 times faster than simulation," Huang said. "With Modulus, scientists will be able to create digital twin models to better understand large systems in unprecedented ways."

Bjorn Stevens, director of the Max Planck Institute for Meteorology, said that improving the resolution of climate models is a key factor in an effective digital twin of the earth. Today's climate models currently rely on statistical workarounds for assessing climate on a global scale, but it is difficult to understand local effects. He said that this is essential for predicting the regional impact of climate change so that we can better inform adaptation efforts.

But Steven suspects that machine learning is some kind of panacea to solve this problem. "There is an illusion that machine learning will replace problems we know how to solve physically, but I think it always has a disadvantage there."

He said that the key to creating a digital twin is to create a highly interactive system, and the beauty of the physical model is that it replicates all aspects of the process in an interpretable way. This is something that a machine learning model trained to mimic the process may not be able to do.

He added that this is not to say that there is no room for machine learning. In areas where we have a lot of data but little knowledge of physics (such as fouling mechanics), it may help speed up workflows, compress data, and potentially develop new models. But he believes that the rapid advancement of supercomputing capabilities means that running physical models at higher resolutions is more of will and resources than capabilities.

The EU hopes to fill this gap with a new program called Destination Earth, which was officially launched in January. The project is a joint effort of the European Space Agency, the European Meteorological Satellite Development Organization and the European Centre for Medium-Term Weather Forecast (ECMWF).

Peter Bauer, deputy director of ECMWF's research department, said that our goal is to create a platform that can bring together various models to simulate key aspects of climate such as the atmosphere and ocean, as well as human systems. "So you have to monitor and simulate not only precipitation and temperature, but also what this means for agriculture, water supply or infrastructure," he said.

Bauer said the result will not be a single homogeneous simulation of all aspects of the earth, but an interactive platform that allows users to introduce any necessary models and data to answer the questions they are interested in.

The project will be implemented gradually over the next ten years, but the first two digital twins they hope to deliver will include one designed to predict extreme weather events such as floods and forest fires, and the other designed to provide long-term forecasts to support climate adaptation and mitigation efforts.

Although Nvidia’s announcement of a new supercomputer dedicated to climate modeling is popular, Bauer said today’s challenges are more about software engineering than developing new hardware. Most key models are developed separately using very different methods, so letting them communicate with each other and find a way to connect data streams of different heights is a prominent problem.

"Part of the challenge of really hiding the diversity and complexity of these components away from users and making them work together," Bauer said.