Archive
All posts
Everything, most recent first. 167 entries so far.
-
Jul 6, 2026Machine LearningWhy plain self-attention chokes on a million-point mesh (quadratic cost), and how Transolver's Physics-Attention fixes it: softly assign each mesh point to one of a small fixed ...
-
Jul 6, 2026Machine LearningHow DoMINO (NVIDIA's Decomposable Multi-scale Iterative Neural Operator) predicts aerodynamics directly from a car's STL/point cloud. It encodes the whole geometry into a global...
-
Jul 6, 2026Machine LearningWhy a plain 3D convnet on a car mesh is cubic (O(N^3)) and infeasible at high resolution, and how FIGConvUNet gets a global receptive field at quadratic (O(N^2)) cost: factorize...
-
Jul 6, 2026Machine LearningPart I of a physics-ML monograph series on NVIDIA's simulation stack for vehicle engineering. What the PhysicsNeMo framework is, how its four repositories fit together, and a cu...
-
Jul 6, 2026Machine LearningPart II of the physics-ML series. What it means for a simulator to have gradients, how NVIDIA Warp turns plain Python into GPU-speed physics code, how the Newton engine builds a...
-
Jul 6, 2026Machine LearningPart III of the physics-ML series. Part I learned the physics; Part II differentiated it — this part is about everything around the physics: the PhysX engine that runs productio...
-
Jul 6, 2026Control SystemsA plain-language field guide to Design and Process FMEA. Imagine how a design or process can fail, decide which failures matter, and fix the worst before a customer ever meets t...
-
Jul 6, 2026Control SystemsHow a vehicle decides something is wrong, writes it down, lights a lamp, protects itself, and how that same machinery becomes the foundation for predicting failures before they ...
-
Jul 6, 2026Machine LearningA reading map for the time-series field guide. One arc told across the series: from a single sensor sample, to a calibrated forecast of its next value, to a forecast of the fail...
-
Jul 6, 2026Machine LearningPart six of the time-series field guide: structure built in, structure learned. Two complementary ways to give a time-series model an edge — physics-informed models that bake th...
-
Jul 6, 2026Machine LearningPart five of the time-series field guide: when the features learn themselves. The neural architectures that learn time-series structure straight from data — recurrent networks (...
-
Jul 6, 2026Machine LearningPart four of the time-series field guide: keeping watch after deployment. A model that was accurate at launch quietly rots as the data drifts, sensors break, and the world moves...
-
Jul 6, 2026Machine LearningPart three of the time-series field guide. Detecting that a machine is sick is diagnostics; saying how many cycles it has left, and how sure you are, is prognostics. A plain-lan...
-
Jul 6, 2026Machine LearningPart two of the time-series field guide: from the past to the next value. How to turn a history of samples into a calibrated forecast of what comes next — moving averages and ex...
-
Jul 6, 2026Machine LearningPart one of the time-series field guide. Every sensor is a storyteller that can only say numbers, one at a time, forever. A plain-language, first-principles tour of how those ra...
-
Jul 6, 2026Stock MarketVolume One of a from-scratch stock-market series, and the entry point to the rest. The market speaks its own language — bullish, bearish, spreads, stop losses, dead cat bounces ...
-
Jul 6, 2026Stock MarketVolume Two. Order books, earnings, the Fed, fear and greed: the forces that tilt every chart you will ever look at, explained from first principles with figures you can push on.
-
Jul 6, 2026Stock MarketVolume Three. A quiz, pattern flashcards, judgment drills, and a trade journal — everything from the earlier volumes, turned into reps. Wrong answers here cost nothing.
-
Jul 6, 2026Stock MarketVolume Four. The income statement, balance sheet, and cash flow, made visual: five questions that tell you in minutes whether a business is healthy, and the red flags hiding in ...
-
Jul 6, 2026Stock MarketVolume Five. Margin, short selling, and options, explained honestly: how they work, why they destroy beginners, and the rules for the day, far from now, when you might consider ...
-
Jul 6, 2026Stock MarketVolume Six. Allocation, the three-fund design, rebalancing, dividend compounding, and taxes: the volume about the next thirty years rather than the next thirty minutes.
-
Jul 6, 2026Stock MarketVolume Seven. Every rule, number, and red flag from all the volumes, compressed onto one screenshot-friendly page.
-
Jul 6, 2026Stock MarketVolume Eight. Microsoft, July 2026: a strong business, a falling stock, and experts disagreeing on the same screen. Every tool in the series, applied to one real research page, ...
-
Jul 6, 2026Stock MarketVolume Nine. Fourteen concepts from a real research screen, each explained in plain words with its own simulation: multiples, PEG, forward earnings, capex J-curves, sum of the p...
-
Jul 6, 2026Stock MarketVolume Ten. Who is on the other side of your trade, what panic and euphoria look like on the tape, why selling mechanically begets selling, and how a calm participant acts in ev...
-
Jul 6, 2026Stock MarketVolume Eleven. Why market crashes are structurally guaranteed, the symptoms that mark a fragile market, the eight phases every crash walks through, and exactly what to do before...
-
Jul 6, 2026Stock MarketVolume Twelve. Whether stock prices can be predicted, honestly answered: nearly impossible tomorrow, hard next quarter, and surprisingly arithmetic over a decade. One method per...
-
Jul 5, 2026Neural NetworksWhere neural operators fit vehicle health: as fast surrogates for the expensive physics simulations — battery thermal fields, electrochemical state, structural response — that a...
-
Jul 5, 2026Neural NetworksNeural operators learn mappings between functions, not just numbers — so one trained model can solve a whole family of physics problems at once, and generalize across resolution...
-
Jul 4, 2026Neural NetworksWhere graph neural networks fit vehicle health monitoring. A battery pack of cells, a powertrain of coupled components, and a fleet of related vehicles are all graphs — and lear...
-
Jul 3, 2026Neural NetworksA single, complete introduction to graph neural networks, in two parts. Part 1 (explained simply) walks the anatomy from the ground up — the input encoder, the stack of message-...
-
Jul 2, 2026Neural NetworksA reading map for the estimation-and-learning thread — physics-informed neural networks, classical Kalman and particle filters, and diffusion models — and how they connect for e...
-
Jul 1, 2026Neural NetworksHow diffusion models turn noise into structure. Add noise to data until nothing is left, then train a network to undo it one small step at a time; sampling runs that reversal to...
-
Jul 1, 2026Neural NetworksThe classical estimators that track a hidden state from noisy measurements. How the Kalman filter alternates predict and correct to fuse a model with data, why the extended and ...
-
Jun 28, 2026Neural NetworksWhere physics-informed neural networks earn their keep in vehicle health monitoring. Battery state-of-charge and state-of-health estimation, thermal and electrochemical field re...
-
Jun 27, 2026Neural NetworksA single, complete introduction to physics-informed neural networks, in two parts. Part 1 (in plain words) builds the idea from the ground up — the mass-on-a-spring intuition, h...
-
Jun 22, 2026Machine LearningAnomaly detection tells you something is wrong; a classifier tells you which fault it is. In vehicle health monitoring the classifiers that earn their keep are almost all built ...
-
Jun 15, 2026ToolsPart 7 of 7 — the capstone of the Logic and Reasoning series. The question that's been hovering since Part I: what is the difference between *doing* reasoning and *understanding...
-
Jun 15, 2026ToolsPart 6 of 7. The previous five parts mostly treated reasoning as if it happened in a single moment. This part puts time back in. How a reasoning system learns from new evidence....
-
Jun 15, 2026ToolsPart 5 of 7. A reasoning system — a single mind or a society of minds — can be unhealthy in characteristic ways. Bias, confirmation, polarization, echo chambers, conspiracy thin...
-
Jun 15, 2026ToolsPart 4 of 7. Part III treated a multi-agent problem-solver as an instantaneous artifact. This part puts it in time. What changes when the team has a history, when it has institu...
-
Jun 14, 2026ToolsPart 3 of 7. Single-agent problem-solving from Part II has a ceiling. Multi-agent problem-solving — where the work is decomposed across a team of specialised reasoners — is the ...
-
Jun 14, 2026ToolsPart 2 of 7. Part I covered how a human mind reasons. Part II asks the harder question: how much reasoning is *computable* at all? Gödel, Turing, the halting problem, complexity...
-
Jun 14, 2026ToolsPart 1 of a seven-part field guide on logic, reasoning, and how those capacities scale from a single mind to a society of minds — and what happens when that society is made of A...
-
Jun 14, 2026Control SystemsPost 6 of the VHM statistical-methods series — and the one that ties the previous five together. Bayes' rule as the diagnostic framework. The Kalman filter and particle filter a...
-
Jun 14, 2026Control SystemsPost 5 of the VHM statistical-methods series. Real machines don't fail in one signal — they fail across a joint distribution of dozens or hundreds. The classical multivariate to...
-
Jun 13, 2026Control SystemsPost 4 of the VHM statistical-methods series. A fault often shows up not in a single signal but in the relationship between two. The toolkit for measuring relationships is broad...
-
Jun 13, 2026Control SystemsPost 3 of the VHM statistical-methods series. A fault announces itself as a change in time. The skill is deciding, from noisy data, that a change is real and not just scatter, a...
-
Jun 13, 2026Control SystemsPost 2 of the VHM statistical-methods series. The residual — what the model predicted minus what actually happened — is the workhorse signal of model-based diagnosis. This monog...
-
Jun 13, 2026Control SystemsThe classical bedrock of vibration-based condition monitoring. RMS, kurtosis, crest factor in the time domain. The FFT, the power spectral density, order analysis in the frequen...
-
Jun 12, 2026ToolsSome AI doesn't chat. It learns to imitate a physics simulator so closely that it answers in milliseconds what used to take a supercomputer hours. NVIDIA has built a full ladder...
-
Jun 12, 2026ToolsA real-time walkthrough of the agentic RCA framework running on the battery case from Case 1. Every step is shown: what the perceiver sees, what the grounder retrieves, what the...
-
Jun 12, 2026ToolsThe architecture that ties the five-case RCA series together. A human-gated agentic framework that runs the diagnostics-and-prognostics method as a closed loop — perceive, groun...
-
Jun 11, 2026Control SystemsCase 5 of 5 — and the case that almost everyone gets wrong on the first pass. Brake noise and reduced initial bite in a subpopulation of EVs. The Pareto says it's a rotor lot; t...
-
Jun 11, 2026Control SystemsCase 4 of 5. An intermittent HV isolation warning, clustered geographically (coastal climates) and temporally (dawn hours). The Pareto points at a connector lot; the truth is an...
-
Jun 11, 2026Control SystemsCase 3 of 5. Elevated tab-to-busbar weld resistance is creating thermal hotspots in finished battery modules. The Pareto chart points straight at one supplier lot — and is wrong...
-
Jun 11, 2026Control SystemsCase 2 of 5. A tonal whine during battery preconditioning, reported by drivers in cold-climate fleets. The investigation traces three independent design choices — d-axis current...
-
Jun 11, 2026Control SystemsCase 1 of a five-part series on root-cause analysis in vehicle health monitoring. We follow a single anomaly — a cohort of electric vehicles losing battery health faster than th...
-
Jun 11, 2026Reinforcement LearningImitation learning is the part of the machine-learning toolkit closest to how humans actually learn most skills — not from a numerical reward, but from watching someone competen...
-
Jun 10, 2026Machine LearningGeneralization and transfer are not two different topics — they are the same statement about error, separated by a single term: how different the new world is from the old one. ...
-
Jun 9, 2026ToolsPart II of the human-in-the-loop monograph. Where Part I covered foundations and mechanisms, this one picks up the harder problem: learning the human's policy with RL, the advan...
-
Jun 9, 2026ToolsThe most-overloaded phrase in 2026 agentic AI is 'human-in-the-loop' — it gets used to mean everything from 'a human watches a dashboard' to 'a human approves each tool call' to...
-
Jun 9, 2026ToolsAn LLM agent that reasons about a propulsion fault is, by default, doing literary criticism on the symptom — pattern-matching against text it has seen before. A physics-grounded...
-
Jun 8, 2026ToolsMost data-visualization advice in 2026 has the same flaw: it tells you what a good figure looks like and stops short of telling the agent how to produce one. A Copilot skill kit...
-
Jun 8, 2026ToolsA demo reel for a Copilot kit of 26 visual-communication skills — paper-monograph shells, interactive parameter labs, algorithm step players, flowcharts, state-machine animation...
-
Jun 8, 2026ToolsAn AI agent that does literature surveys is only as good as the libraries it knows to search. arXiv for preprints. Semantic Scholar for citation graphs. Crossref for canonical m...
-
Jun 8, 2026Reinforcement LearningReinforcement learning shows up inside modern AI agents in two very different places — pre-deployment as a training method that shapes the base model, and at run-time as the pol...
-
Jun 8, 2026ToolsThe premise is a lens, not a claim: agents do not suffer. But the human nosology of psychiatric disorders turns out to be a remarkably good index of how autonomous systems break...
-
Jun 8, 2026ToolsEmotional intelligence — the ability to read, regulate, and respond to affect — is one of the more interesting capacities to track through the human-to-machine translation. In a...
-
Jun 7, 2026ToolsThe security and governance disciplines around AI are not the same discipline and treating them as one produces gaps in both directions. Security is keeping attackers out and li...
-
Jun 7, 2026ToolsThe same AI tools that have become indispensable for engineers have also become indispensable for attackers. Phishing emails are now grammatically perfect and personalized at sc...
-
Jun 7, 2026ToolsTraining a frontier LLM is a one-time cost, paid by a handful of companies. Inference is the recurring cost, paid by everyone who serves the model — and at scale, inference cost...
-
Jun 7, 2026ToolsEvery developer has shipped an application without really understanding what each component of the data stack does — SQL is the query language, SQLite is the embedded database, ...
-
Jun 7, 2026ToolsThe expanded edition of the multi-agent-systems introduction. Where the short version sketches the territory, this monograph builds it out: the four canonical roles fleshed into...
-
Jun 7, 2026ToolsThe 'one giant agent that does everything' design is a dead end. The pattern that actually works in production is the small team — three or four specialized agents with clear ro...
-
Jun 7, 2026ToolsEvery agentic coding system — Claude Code, GitHub Copilot agents, Cursor, Devin, your own harness — runs the same cycle: gather context, take action, verify the result, repeat. ...
-
Jun 7, 2026Control SystemsThe simplest powerful idea in engineering is the feedback loop: measure what the system is doing, compare it with what you want, push back against the difference, repeat. The sa...
-
Jun 7, 2026ToolsWhen an agent system fails in production, it almost always fails in one of three ways: it loops forever, its plan was wrong from the start, or its tool calls did real damage. Th...
-
Jun 6, 2026ToolsWebGL is the low-level graphics API your browser exposes for talking to the GPU. Three.js is the library that hides 800 lines of WebGL boilerplate behind a 30-line scene-camera-...
-
Jun 6, 2026ToolsThere is a particular shape of disappointment that follows a successful agentic AI demo. The demo runs once, looks magical, and then dies the moment it has to run a hundred time...
-
Jun 6, 2026ToolsVolume I established the problem — cascading errors, super-agent bottleneck, missing verification. Volume II is the build manual: context engineering, tool surface design, orche...
-
Jun 6, 2026ToolsThe agent-framework landscape in 2026 is crowded enough that engineers are now confidently choosing one and being wrong about the trade-offs. Six frameworks have meaningful adop...
-
Jun 6, 2026ToolsThe reason Python won is not that it's the best language at any one thing. It's that it's the second-best language at almost everything, the ecosystem connects across all of tho...
-
Jun 6, 2026ToolsNode.js is the JavaScript runtime that lets the language run outside a browser. npm is the package manager that lets you install other people's JavaScript code as a building blo...
-
Jun 6, 2026ToolsObsidian started life as a Markdown note-taking app and quietly turned into something stranger and more useful: a personal knowledge network. The killer feature is not the edito...
-
Jun 5, 2026Machine LearningPyTorch and TensorFlow are the two frameworks that almost every modern neural network is trained on — but the differences between them, and the reasons one wins in research and ...
-
Jun 5, 2026ToolsThe single most important determinant of an LLM-powered system's quality is not the model. It's what you put in the context window. This monograph walks through the three archit...
-
Jun 5, 2026ToolsHermes Agent is one of the more interesting open-source agent projects to ship in 2026: a self-hosted, autonomous AI built by Nous Research around a closed learning loop — every...
-
Jun 5, 2026ToolsAzure documentation is comprehensive in the same way that an unabridged dictionary is comprehensive — every term is defined, nothing is summarized, and you can spend a month rea...
-
Jun 4, 2026ToolsThe hardest part of learning the LangChain ecosystem is not any single tool — it is figuring out which problem each one solves and where one ends and the next begins. LangChain ...
-
Jun 4, 2026ToolsVS Code Copilot has, over the past two years, quietly turned into the most-used AI development environment outside of dedicated tools like Cursor and Claude Code. It is also one...
-
Jun 3, 2026ToolsAPI, CLI, MCP, ADK — four acronyms that get used as if they were rivals competing for the same job. They aren't. They live at different layers of the same stack, and the fastest...
-
Jun 1, 2026ToolsClaude Code is Anthropic's agentic coding tool — a CLI that reads your codebase, edits files, runs commands, and plugs into your dev environment. It is also one of those tools w...
-
Jun 1, 2026ToolsOpenClaw — formerly Clawdbot — is one of the more interesting open-source AI agent projects to appear in the past year: a long-running process that runs on your own hardware, sp...
-
Jun 1, 2026Neural NetworksAttention is the one mechanism that pulled neural networks out of recurrent dead-ends and into the modern era. Every modern language model — GPT, Claude, Gemini, Llama — is some...
-
May 31, 2026ToolsYou pick the same model in VS Code that you use in the Claude app, ask a similar question, and the editor's answer feels thinner. The usual conclusion — 'the IDE version must be...
-
May 31, 2026ToolsArchitecture decisions are the ones you make in the first week of a project and then live with for years. This primer is the working engineer's introduction — what architecture ...
-
May 30, 2026ToolsAn LLM reads text. Source code is text. So why isn't this trivial? Because raw .c files come with a preprocessor that has to be expanded before the code is meaningful, with #inc...
-
May 29, 2026ToolsA 'skill' is the most underrated primitive in agent design: a folder of files — a SKILL.md, some examples, maybe a script — that tells the agent how to do one thing well. The fo...
-
May 28, 2026ToolsThere are two different questions you can ask of any LLM-powered system: where did this answer come from, and how long will the model remember it. The first is about knowledge —...
-
May 27, 2026ToolsThe companion to 'How LLMs See an Image.' Audio is more like text than you'd expect — a 1-D signal turned into a 2-D spectrogram, sliced into patches, and tokenized. Video is au...
-
May 26, 2026ToolsThere is no version of GPT-4o, Claude, or Gemini that 'looks at' an image the way you do. The model is a transformer; it reads tokens. So every image you upload is quietly turne...
-
May 25, 2026Machine LearningFor roughly a decade, XGBoost was the single algorithm most likely to win a Kaggle competition that did not involve images or natural language. It is still, on tabular data, the...
-
May 24, 2026Motor ControlEvery modern electric vehicle is built around three electrical circuits that interact constantly but operate at very different voltages, currents, and timescales. This primer is...
-
May 23, 2026ToolsCAN was designed by Bosch for early-1990s passenger cars and has, against all expectations, outlasted nearly every networking technology of its era. It runs the entire embedded ...
-
May 22, 2026ToolsAI-assisted coding has changed the daily practice of engineering in ways the older how-to books haven't caught up with. This compendium is the long version of what I've learned ...
-
May 21, 2026Neural NetworksAn interactive walkthrough of the thirteen ideas every neural-network practitioner reaches for — dataset splits, signal preprocessing, the bias–variance trade-off, gradient desc...
-
May 20, 2026Neural NetworksThe long-form reference behind the interactive neural-networks primer. Eighteen entries — train/val/test splits, preprocessing pipelines, Savitzky–Golay smoothing, outlier metho...
-
May 19, 2026Control SystemsA seven-act visual walkthrough of multi-agent control — from the simplest consensus problem (six agents averaging with their neighbors) to fully-heterogeneous containment, where...
-
May 18, 2026Control SystemsNearly every dynamical system worth controlling is nonlinear. Nearly every tool we have — stability proofs, frequency analysis, MPC — is built for linear systems. Bernard Koopma...
-
May 17, 2026Control SystemsContinuous-time control theory is taught in s-domain. Real control runs at a fixed sample rate on a microcontroller, in z-domain, with finite numerical precision and a one-sampl...
-
May 15, 2026Control SystemsA fixed-gain controller is a promise that the plant you tuned against is the plant you'll meet in the field. When that promise breaks — drift, wear, operating-point change, unit...
-
May 14, 2026Machine LearningEvery model in machine learning, every system identification routine, every adaptive controller has an optimization problem at its core. This reference walks through the algorit...
-
May 13, 2026Control SystemsOptimal control gets taught as proofs and Riccati equations; in practice it lives or dies on tuning, discretization, conditioning, and whether your solver finishes in the sample...
-
May 6, 2026RoboticsA visual primer on the three localization-and-mapping problems every mobile robot has to solve — pure localization, pure mapping, and the joint estimation problem (SLAM). Eight ...
-
May 6, 2026RoboticsA visual primer on robot motion control — from path tracking on wheeled vehicles (pure pursuit, Stanley, MPC), through arm navigation (inverse kinematics, RRT-Connect), to aeria...
-
May 5, 2026ToolsA field report from setting up an NVIDIA Jetson Orin Nano Super for AI development. NVMe migration, swap on a 1TB drive, GPU-aware Docker, MAXN power mode, and the small handful...
-
May 4, 2026ToolsA 45-minute walkthrough for engineers who keep needing the terminal but never quite sat down to learn it. Five parts: what Linux even is, the file system, the essential commands...
-
May 3, 2026Control SystemsA field manual on twelve practical techniques most control textbooks under-treat: Smith predictors, 2-DOF PID, anti-windup, feedforward, derivative filtering, cascade, gain sche...
-
May 2, 2026Machine LearningAn interactive reference on regression — predicting continuous numbers from data. Nine algorithm families compared, four live demos, and an honest treatment of the data and metr...
-
May 2, 2026RoboticsA primer on self-awareness — the quiet but startling fact that you know there is a you — and how the question is being taken up by roboticists trying to give machines a self-mod...
-
May 1, 2026Control SystemsAn interactive notebook on the structural properties of linear state-space systems — controllability, observability, the Kalman decomposition, Luenberger observers, and Kalman f...
-
Apr 30, 2026Machine LearningAn illustrated essay on support vector machines and their one-class variant — the maximum-margin hyperplane, soft margins, the kernel trick, and the surprising move that turns a...
-
Apr 30, 2026Control SystemsAn interactive tour of classical compensator design. Pick a controller (P / PI / PD / PID / lead / lag / lead-lag), pick a plant, drag the parameters, and watch the Bode plot an...
-
Apr 29, 2026Machine LearningA field guide to the six paradigms of machine learning — supervised, unsupervised, reinforcement, imitation, transfer, and the rapidly-merging boundaries between them. The mecha...
-
Apr 28, 2026RoboticsHow a robot, a car, or a vacuum cleaner decides where to go and which way to get there — the principal families of path planning algorithms, with interactive figures you can tou...
-
Apr 27, 2026Control SystemsAn interactive primer on system identification — the practical art of getting a working model of a real plant. Step tests, PRBS, transfer-function and state-space fitting, valid...
-
Apr 26, 2026Control SystemsTen ideas that turn up almost every time a real controller meets a real plant — feedforward, time delay, Padé approximants, non-minimum phase, gain scheduling, the step response...
-
Apr 25, 2026Control SystemsA field guide to filters — what they are, why classical electrical engineering spent decades designing them, and the seven chapters that take you from RC pairs through the bilin...
-
Apr 24, 2026Control SystemsAn interactive primer on automotive diagnostics — UDS, the DTC structure, OBD, sessions and security, and what changes for the electrified powertrain. The protocol underneath ev...
-
Apr 23, 2026RoboticsAn interactive primer on what a robot actually is — the geometry of reach, the closed-loop control underneath every motion, the sensing that lets it know what it's doing, and th...
-
Apr 22, 2026Machine LearningA 22-chapter field guide to modern AI systems — tokenization, attention, LLMs, prompting techniques, RAG, agent frameworks, MCP, and the engineering considerations around them.
-
Apr 21, 2026Control SystemsMonograph No. 4 of the diagnostics series. The gap between textbook FDI and deployed FDI — domain shift, certification, edge-deployment, fleet-scale digital twins, unlabeled ano...
-
Apr 20, 2026Reinforcement LearningThe most pragmatic post in the field guide — the hyperparameters that almost always work, the order to debug a policy that won't learn, and the silent failures that cost days be...
-
Apr 20, 2026Reinforcement LearningAlgorithms are maybe 30% of a successful RL robotics project. The other 70% is engineering: reward design, observation and action spaces, sim-to-real, learning from logged data,...
-
Apr 20, 2026Reinforcement LearningIf you're a control engineer, this is the RL section written for you. Model-based RL learns dynamics and plans against them — 100× more sample-efficient than model-free on real ...
-
Apr 20, 2026Reinforcement LearningEvery RL algorithm is secretly two algorithms sharing a body — one that exploits what it knows, one that explores for what it doesn't. This post covers the exploration methods t...
-
Apr 20, 2026Reinforcement LearningFrom REINFORCE to PPO in one post — the policy gradient theorem, why its variance is ruinous by default, and the three tricks (baselines, critics, GAE) that make it work in prac...
-
Apr 20, 2026Reinforcement LearningThe Bellman equations say what a value function satisfies; temporal-difference learning says how to estimate it from samples. One update rule, two algorithms (SARSA and Q-learni...
-
Apr 20, 2026Reinforcement LearningEvery RL algorithm starts from the same place: write down the MDP. This post covers the formalism a control engineer actually needs — state, action, reward, the discount factor,...
-
Apr 20, 2026Reinforcement LearningA practitioner-oriented field guide to reinforcement learning for control and robotics. Start here: the RL family tree, a five-question algorithm selector, and the rules of thum...
-
Apr 19, 2026Machine LearningA working Bayesian optimization loop you can run in your browser. Watch Expected Improvement, UCB, and PI compete to find the minimum of a deceptive benchmark function — and rea...
-
Apr 19, 2026Machine LearningClick, slide, and watch the posterior update. A working intuition for Gaussian Processes — from the one-sentence definition through the Cholesky math and the honest O(n³) scalin...
-
Apr 17, 2026Control SystemsMonograph No. 3 of the diagnostics series. Remaining Useful Life, degradation modeling, health indicators, the uncertainty question, and the maintenance scheduling that uses all...
-
Apr 15, 2026Control SystemsMonograph No. 2 of the diagnostics series. Passive vs active fault-tolerant control, reference governors, graceful degradation, and the stability question during the transition ...
-
Apr 13, 2026Control SystemsMonograph No. 1 of a four-part series on diagnostics and fault-handling. The classical model-based methods (observers, parity, Kalman), the data-driven methods (residuals, class...
-
Apr 12, 2026ToolsA practical guide for engineers shifting from Claude Desktop to GitHub Copilot in VS Code. The same model, a different room — the three modes, project knowledge, file handling, ...
-
Apr 11, 2026Control SystemsThe unglamorous foundation under the fleet posts. Those posts assumed a clean 'one row per car' table — build attributes, telemetry summary, warranty outcome — and went straight...
-
Apr 11, 2026Control SystemsThe whole diagnosis series, until now, has been offline — an analyst with logs. But the first diagnostician is the car itself. This post drops down to the edge, where an ECU dec...
-
Apr 11, 2026Control SystemsThe diagnosis series found a cause by watching: a charging code (P0AE0) concentrated in cars on firmware ≥ v5.1, pointing to a cold-throttle bug, and a v5.2 patch was proposed. ...
-
Apr 11, 2026Control SystemsThe closing note in the diagnosis series. The earlier posts found the cause; this one asks the harder question — how do you make sure it never happens again? Fixing the affected...
-
Apr 11, 2026Control SystemsThe rigorous companion to Fleet-Scale Root-Cause Analysis. That post showed the logic; this one puts numbers under the figures and closes the gaps. We make 'is this cohort split...
-
Apr 11, 2026Control SystemsThe capstone of the diagnosis set. A single car gives you one observation; a fleet gives you a distribution — and a hypothesis must now explain not just that the code fired, but...
-
Apr 11, 2026Control SystemsThe third post in the diagnosis set. The sufficiency method assumed away three hard realities, and this post tackles them head-on, in plain words. Multiple simultaneous faults: ...
-
Apr 11, 2026Control SystemsThe companion to Root-Cause Analysis Under Uncertain, Heterogeneous Evidence. That post showed how to weigh messy evidence into a cause; this one answers the question it left op...
-
Apr 11, 2026Control SystemsThe methodological companion to the five-case RCA series. Real investigations don't begin with clean synthetic data — they begin with a heap of evidence of unequal trust: a DTC ...
-
Apr 11, 2026ToolsAn interactive primer on CUDA for engineers comfortable with C but new to GPU programming. Twelve demonstrations covering thread hierarchy, memory model, warp execution, memory ...
-
Apr 11, 2026Control SystemsAn index and reading map for the nine-part diagnosis series. It is one long argument: turn untrustworthy evidence into a cause, know when you have enough to conclude, handle the...
-
Apr 11, 2026Control SystemsA plain-English reference for the whole diagnosis series. Around two hundred terms, grouped into thirteen themes and defined in one sentence each: faults and symptoms; the diagn...
-
Apr 10, 2026ToolsTen ideas about the gap between 'works on my laptop' and 'survives in production' — correctness across environments, performance under real load, resilience and operability, and...
-
Apr 9, 2026Control SystemsAn interactive primer on RUL prediction — four algorithm families compared on the same problem. Similarity matching against historical trajectories, ML regression on time-window...
-
Apr 8, 2026ToolsTen ideas about software engineering that have nothing to do with code — what you're actually building, the human surface around it, trust and law and data, and the long tail of...
-
Apr 7, 2026Machine LearningAn illustrated primer on survival analysis. Why time-to-event data is fundamentally different from regression. Censoring, hazard functions, Cox's elegant trick, and the neural e...
-
Apr 6, 2026ToolsA 36-chapter reference monograph covering the working vocabulary of the modern software stack — JSON, APIs, regex, parsers, JSON Schema, ReAct, tokens, embeddings, RAG, BM25, Ta...
-
Apr 5, 2026Machine LearningAn illustrated primer on the Isolation Forest — the anomaly-detection algorithm that flips the usual approach on its head. Five interactive figures showing isolation paths, the ...
-
Apr 4, 2026Machine LearningAn illustrated primer on autoencoders and variational autoencoders. The bottleneck principle, what lives in the latent space, the variational reformulation that turned an unsupe...
-
Apr 3, 2026Control SystemsAn interactive primer on the three classical tools of diagnostic reasoning — fault tree analysis, Bayesian network inference, and ranked hypotheses with priors. Three live figur...
No posts match that search.