Integrating Machine Learning In Engineering Data Analysis

This #paper presents a pathway to tackle the complex process of modernizing building codes and standards, which often struggle to keep up with technological and domain knowledge advancements. This research introduces the concept of "Equivalent Machine Learning Models." In essence, we engineer ML models to comprehend the foundational principles behind codal provisions, effectively learning the rationale and methodology that guided their initial formulation. We cannot rely on the traditional approach to building ML models to realize equivalent models. Thus, the paper also proposes a new methodology in which equivalent models are trained on #data and to analyze the underlying properties of codal provisions in terms of physics principles, engineering intuition, and causal logic. This ensures the equivalent models not only accurately capture the DNA of the provisions but also produce reliable predictions that are inherently #understandable, mirroring the human logic embedded within the original codal provisions. We tested this methodology across seven structural engineering problems documented in several building codes (including the American Society of Civil Engineers, American Concrete Institute, Australian Building Codes Board, and the American Wood Council). These case studies cover #empirical, #statistical, and #theoretical codal provisions. Our findings indicate: 1. Equivalent Machine Learning Models have the potential to be easily integrated into future building codes, offering a faster, more efficient path to adoption. 2. Despite achieving high predictive accuracy, the “traditional” approach to building ML models is likely to suffer in capturing the properties of codal provisions. 3. This is one tiny step, and more work is evidently needed. Your thoughts and feedback could further pave the way for innovative solutions in our field. As always, I am grateful for the forward-thinking perspective and insights provided by the reviewers and editors of the ASCE’s Journal of Structural Engineering & the SEI - Structural Engineering Institute Paper: https://lnkd.in/e5EfqhKt Preprint on ResearchGate: https://lnkd.in/eJ4Xzruz #Construction #Civilengineering #buildingcodes #machinelearning 

🔬 When Finite Element Modeling Meets Machine Learning in Structural Engineering 🏅 A recent study examines how combining physics-based simulation with machine learning can improve prediction of advanced composite column behavior. The focus is on FRP-confined double-skin tubular columns (DSTCs) — structural members composed of: an outer fiber-reinforced polymer (FRP) tube, an inner steel tube, and a concrete core between them. ✨ This hybrid configuration has attracted interest because it can provide high strength, corrosion resistance, and improved confinement compared with conventional systems. However, predicting their axial behavior is challenging. The interaction between concrete, steel, and FRP introduces nonlinear responses that are difficult to capture using experiments alone. 🧠 Physics-Based Modeling + Data-Driven Prediction The study combines two approaches: 1️⃣ Finite element modeling, validated against experimental results, to simulate structural behavior under axial loading. 2️⃣ #MachineLearning models, trained using both experimental data and #FEM-generated results, to predict ultimate load capacity and axial strain. Several machine learning methods were evaluated, with ensemble models and hybrid approaches showing strong predictive performance for the dataset considered. Importantly, the machine learning models are not used as replacements for mechanics-based analysis, but as tools to accelerate prediction once reliable simulation and experimental data are available. 🏗️ Engineering Insights Concrete filling inside the inner steel tube increases axial capacity and deformation capacity compared with hollow configurations. FRP confinement stiffness and thickness significantly influence column performance. Material and geometric parameters interact strongly, reinforcing the need for integrated modeling approaches. 🚧 Why This Matters As structural systems incorporate more composite materials, design space exploration becomes increasingly complex. Combining validated numerical models with data-driven prediction offers a way to evaluate many design scenarios efficiently while remaining grounded in structural mechanics. For students, this work also illustrates an important shift in engineering practice: AI is not replacing mechanics — it is becoming a tool that extends what mechanics-based models can do. 📄FREE download of the full-text: https://lnkd.in/eT9fUyti #DoubleSkinTubularColumns #FRP #FiberReinforcedPolymer #FiniteElementAnalysis #HighStrengthConcrete #ML #JIPR #newPub #CivilEngineering #StructuralEngineering

🔬 Building ML-Assisted Experimental Ecosystems: A Bottom-up Approach Creating machine learning (ML)-enabled experimental ecosystems is incredibly complex. With countless possible connections and decision-making processes, foundational models alone can’t simply “give us answers.” So, where do we begin? Step 1: Accelerate Data Acquisition Start by developing ML workflows that enhance data collection efficiency on a single instrument. Here, sample selection and data interpretation remain as in traditional setups, but by identifying internal efficiencies, we’re able to operate faster without changing core processes. Step 2: Build Upstream Feedback Next, integrate characterization results to inform upstream sample selection. This creates a feedback loop, refining experiment planning by better aligning initial sample choices with desired outcomes—an early step toward smarter, data-driven experiment planning. Step 3: Enhance Downstream Data Analytics Finally, improve downstream analysis by updating theoretical models based on new data, ultimately generating knowledge. This strengthens our ability to interpret results in ways that update and refine our scientific understanding. But is this enough? Not quite. In reality, we deal with multiple instruments, researchers, and planning decisions that can be connected in workflows in a vast number of combinations. Designing these connections is a challenge in itself and could benefit from approaches in auction theory, game theory, or other complex decision-making frameworks. However, the first step is to build connections that allow all these elements to exist within a shared knowledge space. #MachineLearning #ExperimentAutomation #ScientificWorkflows #AIforScience #MaterialsScience #ResearchInnovation

AI/ML for Engineers – Learning Pathway, Part 2 (Datasets, Code, Projects & Libraries for CAE & Simulation) If you're a mechanical or aerospace engineer diving into ML, you’ve probably realized this: There's no shortage of ML tutorials but very few tailored to simulation, CFD, or physics-based modeling. This second part of Justin Hodges, PhD's blog fills that gap. In the blog, you will find: ➡️ Which datasets actually matter in CAE applications. ➡️ Beginner-friendly vs. advanced datasets for meaningful projects. Links to real engineering data like: ➡️ AhmedML, WindsorML, DrivaerML (31TB of aero simulation data) ➡️ NASA Turbulence Modeling Challenge Cases (with goals for ML-based prediction) ➡️ Johns Hopkins Turbulence Databases ➡️ Stanford CTR DNS datasets, MegaFlow2D, Vreman Research, and more He also points to coding libraries, open-source projects, and suggestions for portfolio-building Especially helpful if you're not publishing papers or attending conferences. Read the full blog here: https://lnkd.in/ggT72HiC Image Source: A Python learning roadmap suggested by Maksym Kalaidov 🇺🇦 in CAE applications! He is a great expert to follow in the space of ML surrogates for engineering simulation. #mechanical #aerospace #automotive #cfd #machinelearning #datascience #ai #ml

Sharing our new study with Hazhir and Ali, where we used machine learning to improve simulation models—including system dynamics models. We applied Neural Posterior Estimation, training neural networks on synthetic data to efficiently estimate model parameters and their uncertainties. Tested on a simple Random Walk model and a more applied pandemic model, this likelihood-free inference approach shows how data and models can be better integrated to address complex systems!

How Mechanical and Materials Engineers Can Start Using AI in Their Work Artificial Intelligence is no longer limited to computer science, it’s becoming an essential tool across disciplines, including engineering and academic research. For mechanical engineers, materials scientists, and educators, here are some practical ways to begin integrating AI into your workflow: 1. Automated Literature Reviews Tools like Elicit, Connected Papers, and ResearchRabbit use AI to identify relevant studies, suggest related work, and even generate summaries; saving hours of manual searching. 2. Data Analysis and Visualization AI-integrated platforms (e.g., PandasAI, ChatGPT Code Interpreter) can help analyze experimental data such as stress-strain curves, thermal profiles, or SEM image results. This can be particularly useful for high-throughput testing or large datasets. 3. Assistance with Simulations For those working with FEA or thermodynamic modeling (e.g., using COMSOL, ANSYS, or CALPHAD), AI tools can help debug code, suggest boundary conditions, or optimize parameters more efficiently. 4. AI in Teaching and Assessment Educators can use AI to generate quizzes, explain complex topics in simpler terms, and even provide feedback on written assignments. It can also support personalized learning pathways for students. 5. AI for Research Planning GPT-based tools can assist with writing research proposals, identifying potential research gaps, and even outlining experimental plans. 6. Exploring AI-Driven Design Algorithms like genetic algorithms, reinforcement learning, or neural networks can be trained to assist in materials discovery, structural optimization, or predictive modeling. Getting Started: • Choose one task from your current workflow (e.g., paper summary, data cleaning, teaching content creation). • Use a trusted AI tool to assist and not replace the process. • Evaluate and refine your use of the tool based on outcomes. AI is not a replacement for engineering knowledge; it’s a powerful extension of it. If you’re already using AI in your work, what tools have been most helpful to you? #AIinEngineering #MechanicalEngineering #MaterialsScience #AcademicResearch #EdTech #CALPHAD #FEA #PhDLife

From Raw Engineering Data to Real-Time Optimization - How is an AI model built? 🔍 This visual from Narnia Labs breaks it down beautifully, and it’s not just a generic ML pipeline. It’s specifically designed for engineering data like 3D CAD models, simulation outputs, and performance metrics. What stood out to me: → Engineering-first data pipeline: It all starts with 1D/2D/3D design or simulation data - no text or tabular shortcuts here. → AI without parameter chaos: Instead of manually defining design variables, generative models learn directly from shapes and performance outcomes. →From model to action: Once trained and tested, the AI can be deployed as an API or GUI for real-time design evaluation and optimization. 🔍 This approach is grounded in a recent peer-reviewed paper from KAIST/Narnia Labs exploring eight application scenarios for generative AI in engineering - from 3D shape generation to simulation prediction and optimization. Instead of optimizing in a high-dimensional parameter space, they compress the design into a low-dimensional latent space - which makes real-time generative optimization possible. That’s a game-changer for anyone working on simulation-heavy or geometry-rich products. 📄 Full paper: Generative AI-driven Design Optimization (Kang, JMSTA 2025) || 🔗 DOI: 10.1007/s42791-025-00097-1 #engineering #ai #generativedesign

I have a master’s in computer science and 13+ years working in analytics. I was shocked when I realized this: Most real-world data science isn’t about Gen AI or deep neural networks. It’s about profiling your data and engineering effective features. Here are 5 reasons why this matters: 1) Decision tree ML is king. Machine learning algorithms based on decision trees are the standard in real-world business analytics. Why? Because they are remarkably effective when your data comes in tabular form. You know, most real-world data. 2) Data are ML's raw materials. Garbage in, garbage out (GIGO) 100% applies to machine learning. However, GIGO is nuanced when it comes to machine learning. For example, many ML algorithms can't handle missing data. While some (e.g., decision trees) can. Which means... 3) Thou shalt profile your data. Profiling your data gives you insight into your raw materials. Here are some examples: 1 - Missing values? 2 - Rare categorical values? 3 - Uniform distributions? 4 - Outlier data-time values? The list goes on. 4) The best models are born from the best features. Decision tree models can learn many things from your data. For example, they will automagically learn feature interactions. However, they can't learn everything. That's where your knowledge of the data is invaluable to... 5) Engineer features. Here's your superpower: Combining your knowledge decision trees with your business process knowledge. This is how you brainstorm and test features to arrive at the best ML models. Remember - process knowledge makes for the best models! 📌 If you're ready to build DIY data science skills, I can help. I send ML tutorials each week to 6,975 professionals. These professionals are also learning: Python Logistic regression K-means cluster analysis Decision tree machine learning With my free crash courses: https://lnkd.in/e7fVrjxC

Modern manufacturing excellence requires seamless integration of machine learning operations (MLOps) within converged IT/OT environments, creating the foundation for true Industrial DataOps. This structured approach enables organizations to deploy, monitor, and continuously improve AI models while maintaining data integrity. Three 🔑 core capabilities manufacturers must have: 1️⃣ Continuous Model Evolution: MLOps pipelines automatically retrain models as production conditions change, maintaining detection accuracy and preventing model drift that would otherwise lead to increased false positives or missed quality issues. 2️⃣ Cross-Disciplinary Collaboration: Standardized governance frameworks like Unity Catalog create common ground where data scientists, IT specialists, and manufacturing engineers can jointly develop, test, and deploy AI solutions that respect operational constraints while leveraging enterprise data resources. 3️⃣ Scalable System Architecture: A properly implemented MLOps strategy enables organizations to scale successful AI implementations from pilot projects to enterprise-wide deployments, replicating proven models across multiple facilities while preserving crucial site-specific customizations. #IndustrialAI #AI #Governance

This Siemens white paper explains how AI-assisted Modal Analysis in Simceter Testlab helps engineering teams handle increasing test complexity, large data sets, and tight timelines. By combining Machine Learning with established Modal Analysis methods in Simcenter Testlab, engineers can reduce manual effort, improve consistency, and generate reliable modal models faster. A practical look at how AI is enhancing,not replacing, engineering expertise in structural dynamics testing. 👉 https://lnkd.in/gqMCY-Qb #Simcenter #Testlab ##ModalAnalysis