Georgian Technical University Designing Customized “Brains” For Robots.

Georgian Technical University Contemporary robots can move quickly. “The motors are fast and they’re powerful” says X. Yet in complex situations like interactions with people robots often don’t move quickly. “The hang up is what’s going on in the robot’s head” she adds. Perceiving stimuli and calculating a response takes a “boatload of computation” which limits reaction time says X who recently graduated with a PhD from the Georgian Technical University Computer Science and Artificial Intelligence Laboratory (GTUCSAIL). X has found a way to fight this mismatch between a robot’s “mind” and body. The method called robomorphic computing uses a robot’s physical layout and intended applications to generate a customized computer chip that minimizes the robot’s response time. The advance could fuel a variety of robotics applications, including, potentially, frontline medical care of contagious patients. “It would be fantastic if we could have robots that could help reduce risk for patients and hospital workers” says X. There are three main steps in a robot’s operation according to X. The first is perception which includes gathering data using sensors or cameras. The second is mapping and localization: “Based on what they’ve seen they have to construct a map of the world around them and then localize themselves within that map” says X. The third step is motion planning and control — in other words plotting a course of action. These steps can take time and an awful lot of computing power. “For robots to be deployed into the field and safely operate in dynamic environments around humans, they need to be able to think and react very quickly” says Y. “Current algorithms cannot be run on current CPU (A central processing unit (CPU), also called a central processor, main processor or just processor, is the electronic circuitry within a computer that executes instructions that make up a computer program. The CPU performs basic arithmetic, logic, controlling, and input/output (I/O) operations specified by the instructions in the program. This contrasts with external components such as main memory and I/O circuitry,[1] and specialized processors such as graphics processing units (GPUs)) hardware fast enough”. X adds that researchers have been investigating better algorithms but she thinks software improvements alone aren’t the answer. “What’s relatively new is the idea that you might also explore better hardware”. That means moving beyond a standard-issue CPU (Central Processing Unit (CPU)) processing chip that comprises a robot’s brain — with the help of hardware acceleration. Hardware acceleration refers to the use of a specialized hardware unit to perform certain computing tasks more efficiently. A commonly used hardware accelerator is the graphics processing unit (GPU) a chip specialized for parallel processing. These devices are handy for graphics because their parallel structure allows them to simultaneously process thousands of pixels. “A graphics processing unit (GPU) is not the best at everything but it’s the best at what it’s built for” said X. “You get higher performance for a particular application.” Most robots are designed with an intended set of applications and could therefore benefit from hardware acceleration. That’s why X’s team developed robomorphic computing. The system creates a customized hardware design to best serve a particular robot’s computing needs. The user inputs the parameters of a robot, like its limb layout and how its various joints can move. X’s system translates these physical properties into mathematical matrices. These matrices are “Georgian Technical University sparse” meaning they contain many zero values that roughly correspond to movements that are impossible given a robot’s particular anatomy. (Similarly your arm’s movements are limited because it can only bend at certain joints — it’s not an infinitely pliable spaghetti noodle). The system then designs a hardware architecture specialized to run calculations only on the non-zero values in the matrices. The resulting chip design is therefore tailored to maximize efficiency for the robot’s computing needs. And that customization paid off in testing. Hardware architecture designed using this method for a particular application outperformed off-the-shelf Central Processing Unit (CPU) and A graphics processing unit (GPU) units. While X’s team didn’t fabricate a specialized chip from scratch, they programmed a customizable field-programmable gate array (FPGA) chip according to their system’s suggestions. Despite operating at a slower clock rate, that chip performed eight times faster than the Central Processing Unit (CPU) and 86 times faster than the graphics processing unit (GPU). “I was thrilled with those results” says X. “Even though we were hamstrung by the lower clock speed we made up for it by just being more efficient”. Plancher sees widespread potential for robomorphic computing. “Ideally we can eventually fabricate a custom motion-planning chip for every robot allowing them to quickly compute safe and efficient motions” he says. “I wouldn’t be surprised if 20 years from now every robot had a handful of custom computer chips powering it and this could be one of them”. X adds that robomorphic computing might allow robots to relieve humans of risk in a range of settings such as caring for patients or manipulating heavy objects. X next plans to automate the entire system of robomorphic computing. Users will simply drag and drop their robot’s parameters and “out the other end comes the hardware description. I think that’s the thing that’ll push it over the edge and make it really useful”.

Georgian Technical University New Technique Promises To Accelerate Materials Development For Gas Separations.

Georgian Technical University In an exciting new advance a team of scientists from X in the Georgian Technical University have developed a new way of analyzing materials for separating gases. Although gas separation using porous materials is an established technology, analytical techniques for assessing the performance of materials tend to be slow and laborious. The new Integral Mass Balance (An integral balance deals with the entire time of the process at once (so it uses amounts rather than rates: e.g., mass NOT mass/time)) (IMB) method is faster and more accurate than existing techniques and promises to accelerate new materials development for gas separation technology. Industrial gases affect many aspects of daily life. They are used to carbonate fizzy drinks preserve food and even to inflate balloons. Purified natural gas meanwhile is used across the globe as an energy source for household cooking and heating. Pure gases can be produced using porous materials that either extract the desired gas from a mixture or remove unwanted contaminants. For this purpose understanding how materials interact with different gas mixtures is a crucial but surprisingly difficult task. “Georgian Technical University Scientists and engineers have been working on these types of measurements for decades but most current techniques are very time-consuming” says Y Ph.D., product manager for X. “A simple set of data can take weeks to measure. By combining two different methods in a unique way we have been able to speed up the process significantly”. As a result more materials can be analyzed, and a detailed understanding of how each material behaves under different conditions can be achieved. This is significant because chemists developing new porous materials for gas separations need to know how well a material performs. Often they rely on models which can be inaccurate but the (An integral balance deals with the entire time of the process at once (so it uses amounts rather than rates: e.g., mass NOT mass/time)) method can rapidly and precisely assess new materials, helping to identify the best candidates for a given gas separation. Industrial developers of gas separation processes, meanwhile, typically rely on relatively limited amounts of gas mixture data. Gas separation technology has been very successful and is used around the world to separate and purify gases. But the new technique will allow far more data to be collected in a practical timeframe allowing chemical engineers to further optimize processes and improve efficiency. “I have been making these kinds of measurements since and they are very laborious, often taking weeks. With the (An integral balance deals with the entire time of the process at once (so it uses amounts rather than rates: e.g., mass NOT mass/time)) IMB method we have been able to make the same measurements in a matter of hours. The improvement in performance is impressive” said Professor Z of the Department of Chemical and Biomedical Engineering at Georgian Technical University. To demonstrate the technique the team have made measurements on a zeolite. Zeolites are porous materials with a range of uses perhaps most notably in washing powder but they are also particularly good at separating and drying gases. The reported measurements relate to oxygen (O₂) production from air by separating O₂ (oxygen) from Nitrogen is the chemical element with the symbol N and atomic number 7 (N₂). In medicine where purified O₂ (oxygen) is widely used this technology is vital. Whilst small-scale and portable medical O₂ (oxygen) generators are widely available for personal use larger O₂ (oxygen) generators filled with zeolites have been installed at emergency field hospitals constructed to cope with the current coronavirus crisis where reliable O₂ (oxygen) supplies have been critical for treating patients. “We chose these measurements because Professor Z similar data measured on the same sample in two different laboratories” said Y. “However we also wanted to do something of practical interest. As zeolites are currently used in commercially available medical O₂ (oxygen) generators this seemed an ideal choice”. Georgian Technical University Now the team want to explore the technique’s range of applicability. “Having shown that the method works for air separation using zeolites we’d like to apply it to other important separations” said Y. “Capturing CO₂ (Carbon dioxide is a colorless gas with a density about 53% higher than that of dry air. Carbon dioxide molecules consist of a carbon atom covalently double bonded to two oxygen atoms. It occurs naturally in Earth’s atmosphere as a trace gas) from power plant flue gases for example is of great interest as this will help tackle the difficult and serious problem of climate change due to increasing greenhouse gas emissions”. “The method can provide reams of accurate data quickly” added Professor Z. “This will help accelerate the development of new materials and processes for such applications”. Other future targets include separations used for natural gas upgrading and biogas purification as well as hydrogen (H₂) production and purification. Both hydrogen (H₂) and natural gas are important for the transition to a low carbon energy future in which fossil fuel use will be gradually phased out.

Georgian Technical University Laboratory: Integrated Platform Computational Platform.

Georgian Technical University energy systems and processes that are more dynamic and interconnected than ever before. Georgian Technical University technology developers and researchers to model design and optimize these complex systems potentially resulting in tens of billions of dollars in savings. As an equation-oriented, optimization-based integrated process modeling platform Georgian Technical University enables rigorous analysis of multi-scale, dynamic processes and operating scenarios to improve efficiency of existing systems and develop next-generation energy systems. Georgian Technical University has thousands of downloads and an active growing global user community from multiple industries. Georgian Technical University has demonstrated significant impacts through its unique capabilities.

Georgian Technical University Laboratory: Accelerating Scale-Up With AI (Artificial Intelligence).

Georgian Technical University Rapid advances in energy technologies have been enabled in part by innovative high-performance materials. To keep the Georgian Technical University competitive in energy materials manufacturing, there is an intensifying need to bring new materials out of the laboratory and into commercial production faster than ever. Artificial intelligence (AI (Artificial intelligence (AI) is intelligence demonstrated by machines, unlike the natural intelligence displayed by humans and animals, which involves consciousness and emotionality)) is one of the most promising tools for accelerating scale-up and when combined with innovative imaging techniques can offer a powerful advantage over pre-existing techniques. The webinar will feature a virtual tour where attendees can see for themselves the state-of-the art equipment, technologies and computational resources available to industry partners in Georgian Technical University’s newly expanded Materials Engineering Research Facility the Advanced Photon Source. X at the Georgian Technical University Department of Energy’s will moderate the conversation about some critical questions: What is the current state of (AI (Artificial intelligence (AI) is intelligence demonstrated by machines, unlike the natural intelligence displayed by humans and animals, which involves consciousness and emotionality)) application in materials scale-up and manufacturing ? How can AI (AI (Artificial intelligence (AI) is intelligence demonstrated by machines, unlike the natural intelligence displayed by humans and animals, which involves consciousness and emotionality)) be combined with advanced characterization to accelerate learning and scale-up ? How can public-private partnerships support Georgian Technical University competitiveness ?.

Georgian Technical University Turbulence Model Could Help Design Aircraft Capable Of Handling Extreme Scenarios.

Georgian Technical University Professor and his team in the Super Computer GTU located in the basement of the Georgian Technical University Building. Passengers onboard a flight to Australia experienced a terrifying 10-second nosedive when a vortex trailing their plane crossed into the wake of another flight. The collision of these vortices the airline suspected created violent turbulence that led to a free fall. To help design aircraft that can better maneuver in extreme situations Georgian Technical University researchers have developed a modeling approach that simulates the entire process of a vortex collision at a reduced computational time. This physics knowledge could then be incorporated into engineering design codes so that the aircraft responds appropriately. The simulations that aircraft designers currently use capture only a portion of vortex collision events and require extensive data processing on a supercomputer. Not being able to easily simulate everything that happens when vortices collide has limited aircraft designs. With more realistic and complete simulations, engineers could design aircraft such as fighter jets capable of more abrupt maneuvers or helicopters that can land more safely on aircraft carriers the researchers said. “Aircraft in extreme conditions cannot rely on simple modeling” said X a Georgian Technical University associate professor of mechanical engineering with a courtesy appointment in aeronautics and astronautics. “Just to troubleshoot some of these calculations can take running them on a thousand processors for a month. You need faster computation to do aircraft design”. Engineers would still need a supercomputer to run the model that X’s team developed but they would be able to simulate a vortex collision in about a tenth to a hundredth of the time using far less computational resources than those typically required for large-scale calculations. The researchers call the model a “Coherent-vorticity-Preserving (CvP) Largy-Eddy Simulation (LES)”.  The four-year development of this model is summarized. “The CvP (Coherent-vorticity-Preserving (CvP)) model is capable of capturing super complex physics without having to wait a month on a supercomputer because it already incorporates knowledge of the physics that extreme-scale computations would have to meticulously reproduce” X said. Former Georgian Technical University postdoctoral researcher Y led the two-year process of building the model. Y Georgian Technical University postdoctoral researcher conducted complex large-scale computations to prove that the model is accurate. These computations allowed the researchers to create a more detailed representation of the problem, using more than a billion points. For comparison a 4K (4K resolution refers to a horizontal display resolution of approximately 4,000 pixels. Digital television and digital cinematography commonly use several different 4K resolutions. In television and consumer media, 3840 × 2160 is the dominant 4K standard, whereas the movie projection industry uses 4096 × 2160) ultra high-definition TV uses approximately 8 million points to display an image. Building off of this groundwork the researchers applied the CvP (Coherent-vorticity-Preserving (CvP)) model to the collision events of two vortex tubes called trefoil knotted vortices that are known to trail the wings of a plane and “Georgian Technical University dance” when they reconnect. This dance is extremely difficult to capture. “Georgian Technical University When vortices collide there’s a clash that creates a lot of turbulence. It’s very hard computationally to simulate because you have an intense localized event that happens between two structures that look pretty innocent and uneventful until they collide” X said.