AI Applications In Engineering

As we transition from traditional task-based automation to 𝗮𝘂𝘁𝗼𝗻𝗼𝗺𝗼𝘂𝘀 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁𝘀, understanding 𝘩𝘰𝘸 an agent cognitively processes its environment is no longer optional — it's strategic. This diagram distills the mental model that underpins every intelligent agent architecture — from LangGraph and CrewAI to RAG-based systems and autonomous multi-agent orchestration. The Workflow at a Glance 1. 𝗣𝗲𝗿𝗰𝗲𝗽𝘁𝗶𝗼𝗻 – The agent observes its environment using sensors or inputs (text, APIs, context, tools). 2. 𝗕𝗿𝗮𝗶𝗻 (𝗥𝗲𝗮𝘀𝗼𝗻𝗶𝗻𝗴 𝗘𝗻𝗴𝗶𝗻𝗲) – It processes observations via a core LLM, enhanced with memory, planning, and retrieval components. 3. 𝗔𝗰𝘁𝗶𝗼𝗻 – It executes a task, invokes a tool, or responds — influencing the environment. 4. 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 (Implicit or Explicit) – Feedback is integrated to improve future decisions.     This feedback loop mirrors principles from: • The 𝗢𝗢𝗗𝗔 𝗹𝗼𝗼𝗽 (Observe–Orient–Decide–Act) • 𝗖𝗼𝗴𝗻𝗶𝘁𝗶𝘃𝗲 𝗮𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲𝘀 used in robotics and AI • 𝗚𝗼𝗮𝗹-𝗰𝗼𝗻𝗱𝗶𝘁𝗶𝗼𝗻𝗲𝗱 𝗿𝗲𝗮𝘀𝗼𝗻𝗶𝗻𝗴 in agent frameworks Most AI applications today are still “reactive.” But agentic AI — autonomous systems that operate continuously and adaptively — requires: • A 𝗰𝗼𝗴𝗻𝗶𝘁𝗶𝘃𝗲 𝗹𝗼𝗼𝗽 for decision-making • Persistent 𝗺𝗲𝗺𝗼𝗿𝘆 and contextual awareness • Tool-use and reasoning across multiple steps • 𝗣𝗹𝗮𝗻𝗻𝗶𝗻𝗴 for dynamic goal completion • The ability to 𝗹𝗲𝗮𝗿𝗻 from experience and feedback    This model helps developers, researchers, and architects 𝗿𝗲𝗮𝘀𝗼𝗻 𝗰𝗹𝗲𝗮𝗿𝗹𝘆 𝗮𝗯𝗼𝘂𝘁 𝘄𝗵𝗲𝗿𝗲 𝘁𝗼 𝗲𝗺𝗯𝗲𝗱 𝗶𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲 — and where things tend to break. Whether you’re building agentic workflows, orchestrating LLM-powered systems, or designing AI-native applications — I hope this framework adds value to your thinking. Let’s elevate the conversation around how AI systems 𝘳𝘦𝘢𝘴𝘰𝘯. Curious to hear how you're modeling cognition in your systems.

The power sector is changing fast, and AI is at the center of this transformation. From predicting outages before they happen to improving energy distribution, AI is making electricity more reliable, efficient, and sustainable. But how exactly is AI reshaping the industry? 1. Predicting failures before they happen. Power outages can be costly and disruptive. AI-powered predictive maintenance helps utilities identify potential failures in transformers, power lines, and substations before they occur. By analyzing data from sensors and historical trends, AI reduces downtime and ensures a more stable power supply. 2. Smarter energy distribution. Electricity demand fluctuates throughout the day. AI helps balance supply and demand in real time, ensuring power is distributed where it’s needed most. This minimizes waste, lowers costs, and improves overall grid efficiency. 3. Optimizing renewable energy. Renewable energy sources like solar and wind are unpredictable. AI helps by analyzing weather patterns and adjusting energy production accordingly. This means more stable integration of renewables into the grid. While AI is transforming the power sector, technology alone isn’t enough. The biggest challenge is adoption. Getting companies, governments, and individuals to embrace these changes. For digital transformation to succeed, the industry needs: → Skilled talent → Better infrastructure → And a willingness to rethink traditional ways of managing power AI is here to stay, and its impact on energy is growing. The question is: Are we ready to maximize its potential?

🚀 Revolutionizing Aerospace Design with Generative AI: The Future of Aircraft Efficiency 🌍✈ In the fast-paced world of aerospace engineering, every gram saved equals more fuel efficiency and less environmental impact. Here’s a game-changing example of how we’re leveraging Fusion 360’s Generative Design to reshape aircraft seat mounting brackets. 💡 The Problem: Traditional aluminum brackets, weighing in at 1,672 grams, are a significant contributor to unnecessary weight and fuel costs. 🌟 The Solution: By incorporating Generative Design, we’ve cut the weight by 54%, reducing the bracket to just 766 grams! 🔍 How it Works: • Topology Optimization: Streamlining material usage while maintaining strength and safety. • Advanced Materials: Magnesium—35% lighter than aluminum—is now a key part of the design. 🛠 Key Benefits: • Weight Reduction: The new design significantly reduces aircraft weight. • Fuel Savings: Less weight = less fuel burned = lower operational costs. • Sustainability: Lighter components contribute to reduced carbon emissions over the long term. • Cost Efficiency: Airlines can potentially save millions in fuel costs across the lifetime of their fleets. 💬 What This Means for the Future of Aerospace: This isn’t just about lighter brackets; it’s about transforming the way we think about efficiency and sustainability in aviation. ✅ Join the conversation: How do you think generative design will impact the future of aerospace engineering? Share your thoughts below! #GenerativeDesign #Fusion360 #AerospaceInnovation #SustainableDesign #FuelEfficiency #AerospaceEngineering #TechForGood #CarbonReduction #AIinEngineering #FuturisticDesign

Traditional Design vs Generative Design – A Shift in Engineering Thinking In the world of mechanical and aerospace engineering, design methods are evolving rapidly. The image above clearly illustrates the contrast between Traditional Design and Generative Design using an example of aircraft seat mounting brackets. 🔹 Traditional Design This approach relies on human intuition, experience, and established standards. Designers use basic geometric shapes and overengineer components to ensure safety, often leading to excess material usage and heavier parts. In the image, the traditional bracket weighs 1,672 grams, made with solid material and a blocky design to ensure strength. However, it lacks material efficiency and may contribute to increased fuel consumption in aircraft. 🔹 Generative Design This is an advanced, AI-driven design process. Engineers input goals (like weight reduction, strength requirements, material type, and load conditions), and the software generates multiple optimized design solutions. The result is often an organic, lattice-like structure that removes unnecessary material. In the image, the generatively designed bracket weighs only 766 grams — a 55% weight reduction — while still meeting performance criteria. 💡 Key Differences: Design Process: Human-driven vs AI-assisted Material Usage: Excessive vs optimized Shape: Simple, blocky vs complex, organic Efficiency: Heavier and stronger than needed vs lightweight and just as strong Generative design is not just a trend—it's a strategic shift toward sustainable, high-performance engineering. It helps industries like aerospace, automotive, and manufacturing to save weight, reduce cost, and innovate faster. This transformation is a perfect example of how technology is redefining the boundaries of what's possible in design and engineering. --- #TraditionalDesign #GenerativeDesign #MechanicalEngineering #CAD #DesignInnovation #AerospaceEngineering #LightweightDesign #TopologyOptimization #FutureOfEngineering #AutodeskFusion360 #EngineeringTransformation #ProductDesign #AIInEngineering

The energy grid is under immense strain from extreme weather, wildfires, and rising electricity demand. As these pressures increase, so does the need for smarter, more resilient and reliable energy grids.   Utilidata, a company that is part of Microsoft's Climate Innovation Fund portfolio, is redefining energy delivery through its AI platform, Karman. This technology empowers utilities to optimize energy delivery and make better decisions about how to manage the grid by, for example, storing electricity in batteries during off-peak hours and distributing it when it's needed. As a result, electric vehicles and solar panels become flexible, valuable assets that help meet grid demand.   Embedding AI directly into the grid infrastructure helps utility decision-makers make more informed decisions and better serve customers. This innovation highlights the power of AI to modernize critical infrastructure and transform the energy sector.

Big breakthrough: A few months my lab at MIT introduced SPARKS, our autonomous scientific discovery model. Since then we have demonstrated applicability to broad problem spaces across domains from proteins, bio-inspired materials to inorganic materials. SPARKS learns by doing, thinks by critiquing itself & creates knowledge through recursive interaction; not just with data, but with the physical & logical consequences of its own ideas. It closes the entire scientific loop - hypothesis generation, data retrieval, coding, simulation, critique, refinement, & detailed manuscript drafting - without prompts, manual tuning, or human oversight. SPARKS is fundamentally different from frontier models. While models like o3-pro and o3 deep research can produce summaries, they stop short of full discovery. SPARKS conducts the entire scientific process autonomously, generating & validating falsifiable hypotheses, interpreting results & refining its approach until a reproducible, fully validated evidence-based discovery emerges. This is the first time we've seen AI discover new science. SPARKS is orders of magnitude more capable than frontier models & even when comparing just the writing, SPARKS still outperforms: in our benchmark evaluation, it scored 1.6× higher than o3-pro and over 2.5× higher than o3 deep research - not because it writes more, but because it writes with purpose, grounded in original, validated compositional reasoning from start to finish. We benchmarked SPARKS on several case studies, where it uncovered two previously unknown protein design rules: 1⃣ Length-dependent mechanical crossover β-sheet-rich peptides outperform α-helices—but only once chains exceed ~80 amino acids. Below that, helices dominate. No prior systematic study had exposed this crossover, leaving protein designers without a quantitative rule for sizing sheet-rich materials. This discovery resolves a long-standing ambiguity in molecular design and provides a principle to guide the structural tuning of biomaterials and protein-based nanodevices based on mechanical strength. 2⃣ A stability “frustration zone” At intermediate lengths (~50- 70 residues) with balanced α/β content, peptide stability becomes highly variable. Sparks mapped this volatile region and explained its cause: competing folding nuclei and exposed edge strands that destabilize structure. This insight pinpoints a failure regime in protein design where instability arises not from randomness, but from well-defined physical constraints, giving designers new levers to avoid brittle configurations or engineer around them. This gives engineers and biologists a roadmap for avoiding stability traps in de novo design - especially when exploring hybrid motifs. Stay tuned for more updates & examples, papers and more details.

Airlines aren’t just talking about AI - they’re already using it to smooth operations, save fuel and keep passengers moving. Delta Air Lines’ Operations Control Centre runs a machine‑learning tool that studies weather patterns and re‑sequences flights hours before storms bite, cutting knock‑on delays. Avionics International easyJet has fitted its entire Airbus fleet with Skywise Predictive Maintenance. Engineers now replace parts before they fail, reducing technical delays and cancellations. Airbus Alaska Airlines dispatchers use Flyways AI to pick the most efficient routes in real time. On long sectors that’s delivering 3‑5 percent fuel and CO₂ savings-over a million gallons a year. Alaska Airlines News PR Newswire Qantas puts personalised fuel‑efficiency analytics in every pilot’s hand via GE’s FlightPulse, driving behaviour changes that trim both fuel burn and emissions. geaerospace.com Lufthansa Systems’ NetLine/Ops ++ aiOCC gives controllers an AI “copilot” that turns masses of live data into recommended actions, helping curb cascading delays across the network. Lufthansa Systems Three take‑aways for carriers still on the fence: AI thrives in the messy middle. It surfaces the next best action when plans unravel. ROI is tangible. Minutes saved, gallons saved, cancellations avoided—every metric lands on the P&L. Humans stay in control. The most successful roll‑outs pair smart algorithms with experienced dispatchers, engineers and pilots. If your airline is still juggling spreadsheets during disruptions, the sky is sending a clear signal: it’s time to bring AI into day‑to‑day ops.

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

AI meets #CFD: 1500+ airfoils, reduced-order models, and deep learning Just discovered a hidden gem for anyone working on aerodynamics, reduced models, or AI-assisted simulation: AI_Airfoil_CFD – an open-source repo that applies CNNs, POD, and FCDNNs to predict aerodynamic performance across 1500+ airfoils from the UIUC dataset. What you’ll find inside: ✅ A full comparison of POD vs. DNN on the Eppler387 airfoil ✅ CNN models trained on UIUC profiles ✅ Preprocessing + visualization scripts ✅ KSCFE course materials and academic references ✅ Perfect for speeding up CFD workflows or building your own digital twin Built by researchers from GIST, this is one of those projects that quietly bridges the gap between simulation engineers and ML engineers. 💡 Repo link: https://lnkd.in/eYmAAug5 #ai #engineering #simulation

40% less cooling energy. 15% drop in PUE. Same hardware. New intelligence. Here’s how you achieve that... step by step... ...not in theory, but in live production halls. Start with the baseline. Without accurate visibility, the AI can’t learn. Instrument everything: • Thermal sensors per aisle • CRAC inlet/outlet delta-T • Chilled water flow, temp, pressure • Rack-level power + temp from PDUs You’re not logging for compliance. You’re training a nervous system. Next: feed the brain. Use that telemetry to train a reinforcement learning model, Google DeepMind-style or via Schneider Electric’s AI Advisor. Don’t overcomplicate this. The agent gets three things: Current state (temps, flows, loads) Control levers (setpoints, fan speeds, valve positions) A reward function (kWh saved, uptime preserved) It simulates thousands of micro-decisions. Penalises inefficiency. Rewards stability. Eventually, it starts predicting how a 0.2°C delta at the chiller outlet will be reflected in rack temperatures 6 minutes later. Now comes closed-loop control. Let the agent write back into your BMS, adjusting in real-time. Fan curves. Water temps. Airflow paths. All tuned dynamically, every few minutes. But always enforce guardrails: • No overrides to safety interlocks • Always keep human supervisory control • Weekly drift audits to recalibrate reward thresholds Validate in a single hall. Watch the numbers, not the dashboard. Look for: • 35–45% reduction in cooling energy • 10–20% drop in site PUE • Reduction in compressor cycling • Improved temp stability at rack inlets Once validated, replicate. Don’t scale chaos... ...scale confidence. In Gulf summers or Delhi heatwaves, static thresholds collapse. Only adaptive systems survive. AI is not your silver bullet. It’s your second brain. Would you trust it to pull the levers? Let’s talk control, not just monitoring.