American Society of Civil Engineers’ Post

I'm glad to share that my master's research was presented at the 41st International Symposium on Automation and Robotics in Construction (ISARC 2024) in Lille, France! That research work was supported by Dr. Ehsan Rezazadeh Azar, PhD., P.Eng. A huge thanks to him for supporting and presenting the work at the conference. Check out the research findings in the article below! #Automation #Construction #machinelearning #deeplearning #ISARC2024 #PMBE

In my days as an engineering undergrad, running CFD calculations often pushed my laptop to its limits, resulting in its early demise. Achieving accurate CFD simulations could take days, depending on the processing power of the computer. As a staunch advocate of AI, I fully approve of this. With the emergence of EV startups and budding race car builders, software solutions like this will provide quick and accurate results, without the need for substantial hardware investments. #CFD #Simulation #ComputationalFluidDynamics #AI #ArtificialIntelligence #Technology #Software #EVstartup #racecars #aftermarket #aerodynamics #Innovation #STEM #EngineeringLife

Throughout the operational lifespan of aircraft, they are subjected to cyclic loading induced by aerodynamic forces. Over time, this leads to fatigue crack initiation and propagation, ultimately compromising the structural integrity of the aircraft. Today marked a significant milestone as I presented my research on "Determining the Extent of Damage in Aircraft Structures Using Deep Learning" at the 13th Annual College of Engineering Poster Competition at North Carolina Agricultural and Technical State University. By seamlessly integrating insights from interdisciplinary fields such as Structural Health Monitoring and Deep Learning, I successfully developed and trained a Conventional Neural Network (CNN) model. This model demonstrates remarkable accuracy in predicting the relative position of cracks to acoustic emission sensors within aircraft structures. This endeavor not only underscores the importance of leveraging cutting-edge technologies but also highlights the critical role of interdisciplinary collaboration in advancing aerospace engineering. I am excited to continue exploring innovative solutions to enhance aircraft safety and reliability in the future. #innovation #deeplearning #nde #ncat

Aerodynamic Results – From Hours to Seconds 🚗 📉 Artificial Intelligence is changing how we solve complex problems. Using Machine Learning, computers can make predictions and decisions based on data. 🌎 Image Source: NAVASTO 📥 Latest newsletter: https://lnkd.in/d7B7fqA #engineeredmind #science #technology #engineering #cfd #aerodynamics

Aerodynamic Results – From Hours to Seconds 🚗📊 Artificial Intelligence is changing how we solve complex problems and computers can make predictions and decisions based on data. What's your first thought when it comes to combining Computational Fluid Dynamics (CFD) with #AI? Follow our page to learn how companies in your industry are accelerating aerodynamics 📈 👉 https://lnkd.in/d9vwqSkC #Automotive #Technology #Engineering #CFD #Aerodynamics

The comment section is really interesting. There are many industries that will be affected by Machine Learning. There are many careers that will be made obsolete with the proliferation and development of such technologies. I have a feeling that there are many people in this comment section who feel this way. Who understand that the career path that they once took, will no longer exist 20 years from now. That the artisan nature of CFD could be lost forever. Am I being dramatic? What are your thoughts on the matter?

AI revolutionizes problem solving in many fields. Yet, let's not forget the critical role of domain expertise in CFD in achieving these rapid results. It's not magic by AI but a dedicated and specialized application of this powerful tool to fully harness its potential in this field.

📢 📰 Paper Alert! "Deep learning with simulated laser scanning data for 3D point cloud classification" . #Preprint: https://lnkd.in/eZDa9PTi Find out how we use models trained on #VirtualLaserScanning (VLS) data to classify real #laserscanning #pointclouds in different geographic domains (urban 🏙️+ leaf-wood 🌳), achieving state-of-the-art results even without real training data involved!   Why VLS?💪 ▶ Large amounts of diverse training data can be generated quickly and at low cost. ▶ Simulation configurations can be adapted so that the VLS training data closely resembles the target real data. ▶ The whole workflow can be automated, e.g. by procedural scene generation.   Further resources: 🔗Blog post: https://lnkd.in/eaHaqvJw 🔗VirtuaLearn3D project: https://lnkd.in/e_2_vxzu 🔗HELIOS++ project: https://lnkd.in/ebpnpVkz 🔗HELIOS++ LiDAR simulation software: https://lnkd.in/gkPSdSEu 🔗HELIOS++ publication: https://lnkd.in/ek4kQWrT   This work was funded by the Deutsche Forschungsgemeinschaft (DFG) - German Research Foundation: #syssifoss and #VirtuaLearn3D) and further institutions. For the complete funding statement, please check the acknowledgements of the paper. Lukas Winiwarter, Hannah Weiser #3DGeo #HELIOS++ #VirtuaLearn3D

How AI helps solve Complex Aerodynamic problems within seconds with repeatability and accuracy

AI revolution in CFD. From 12 hours to 20 milliseconds for a car's simulation is just mind blowing that how fast it can run. Think how fast it can do an aircraft's engine or just the whole aircraft.