Computer Vision Primer
Image brightness and luminance Lab — exposure, gamma and clipping
Compare linear exposure with encoded grey-value adjustments using synthetic images, histograms and gradients.
Run in your browser
Explore conditions & results →LAB / EXPERIMENT & VERIFY
Understand by experimenting.
Change a condition. Compare the results. Verify with code. Make the technology your own, one experiment at a time.
01Choose
Pick a topic and open it in your browser.
02Change
Adjust inputs or control settings.
03Compare
Use plots and numbers to understand why.
Explore experiments
35 Labs
Computer Vision Primer
Feature extraction Lab — compare Harris and Shi–Tomasi
Rotate, resize and change the brightness of synthetic images, then compare detected corners and known reference positions.
Run in your browser
Explore conditions & results →
Computer Vision Primer
Feature matching Lab — descriptor filters, geometric inliers and ground truth
Compare image patches, filter proposed correspondences and see why geometric agreement does not guarantee a correct match.
Run in your browser
Explore conditions & results →
Computer Vision Primer
Optical flow Lab: track motion between two frames
Estimate sparse image motion and compare it with known synthetic motion. Explore brightness changes, occlusion and the aperture problem.
Run in your browser
Explore conditions & results →Control Engineering Primer
MPC Lab — re-solve a curvature sequence within a horizon and steering limits
Change the horizon, weights and steering limit, and compare predicted paths, solver convergence and failure handling in your browser.
Run in your browser
Explore conditions & results →Control Engineering Primer
Path-tracking comparison Lab — run PP, APP, RPP, Stanley and MPC under the same conditions
Run five trackers together on one vehicle, path, initial state and steering limit, and compare cross-track error, smoothness and completion with shared metric definitions.
Run in your browser
Explore conditions & results →Control Engineering Primer
Adaptive Pure Pursuit Lab — scale lookahead with speed
Compare speed-scaled lookahead with PP, including acceleration and minimum/maximum bounds.
Run in your browser
Explore conditions & results →Control Engineering Primer
Pure Pursuit Lab — compare fixed-lookahead path tracking
Change fixed lookahead and inspect tracking error, steering and targets on a shared vehicle model.
Run in your browser
Explore conditions & results →Control Engineering Primer
Regulated Pure Pursuit Lab — speed through curves and near the goal
Separate lookahead from speed regulation and compare curvature limits, speed floors, final approach and actual vehicle speed.
Run in your browser
Explore conditions & results →Control Engineering Primer
Stanley Lab — steering from front-axle cross-track and heading error
Separate front and rear axle errors and compare gain, speed and steering constraints in your browser.
Run in your browser
Explore conditions & results →Controlled Environment Agriculture
CEA sensing and image estimation — sensor placement and projected leaf area
Combine CO₂, EC, pH, DO, flow, level and pressure to separate spatial differences, instrument bias and sample transport delay. Check circulation, ventilation and water-balance residuals.
Run in your browser
Explore conditions & results →
Controlled Environment Agriculture
CEA climate-balance labs — conserve heat, water vapor and CO2 separately
Integrate a synthetic one-zone model in Python and verify how LED heat, outdoor exchange, ventilation, capacitance and cooling change room temperature.
Run in your browser
Explore conditions & results →Controlled Environment Agriculture
CEA climate control — temperature, humidity, sensors, response and coupling
Treat room temperature and vapour mass as two states, then define interactions and test conditions for cooling, dehumidification, reheat and ventilation.
Run in your browser
Explore conditions & results →Controlled Environment Agriculture
CEA plant environment and light design — photosynthesis, quantity, distribution and spectrum
Connect photosynthesis, respiration, stomata, and water transport to environmental measurements. Two synthetic calculations separate carbon gain from harvest mass and vapor pressure deficit from transpiration.
Run in your browser
Explore conditions & results →Controlled Environment Agriculture
VPD and DLI Lab — verify humidity and light time series
Calculate air VPD and daily light integral from synthetic CSV, testing unequal intervals, missing data and unit mistakes.
Run in your browser
Explore conditions & results →
Control Engineering Primer
Run a Kalman filter in Python: noise, missing data and outliers
Compare Q, R and outlier gating on one synthetic trajectory using estimation error and rejected observations.
Run in your browser
Explore conditions & results →
Robotics Primer
Why ICP fails: initialization, outliers and symmetric geometry
Run 2D registration and distinguish residuals, correspondence coverage and true pose error.
Run in your browser
Explore conditions & results →
Notes
Where Docker data lives: layers, volumes and backups
Distinguish writable layers, named volumes and bind mounts, then verify persistence across container recreation.
Run in your browser
Explore conditions & results →
Notes
Measure home-server power: idle, inference and storage
Log wall-side active power consistently and compare idle demand and energy per completed workload.
Run in your browser
Explore conditions & results →
Machine Learning Primer
Benchmark local LLMs: first response, generation rate and memory
Measure first answer content separately from generation rate and control residency, caching and memory conditions.
Run in your browser
Explore conditions & results →
Powertrain & Mechanical Engineering
Robot reducers: planetary, strain-wave and cycloidal mechanisms
Compare three mechanisms and connect speed, torque and inertia with selection criteria beyond backlash.
Run in your browser
Explore conditions & results →
Power Electronics Primer
From motor speed control to FOC: current, torque and nested loops
Understand how speed control requests current and how FOC connects rotating coordinates, electrical angle and limits.
Run in your browser
Explore conditions & results →
Robotics Primer
Nav2 has a path but will not move: a diagnostic workflow
Inspect geometry, costmaps, control feasibility and command delivery to explain a stopped robot.
Run in your browser
Explore conditions & results →
Power Engineering Primer
Simulate grid frequency: imbalance, inertia and primary response
Use a simple one-area model to separate inertia from primary frequency response after a supply deficit.
Run in your browser
Explore conditions & results →
Robotics Primer
Robot coordinate transforms: matrices, quaternions and TF
Use a numeric example to connect rotation, translation, inverse transforms, quaternions and ROS frames.
Run in your browser
Explore conditions & results →
Robotics Primer
Reproduce robot failures with rosbag2: inputs, TF and time
Build controlled replay comparisons and diagnose duplicate TF publishers and mixed time sources.
Run in your browser
Explore conditions & results →
Sensors & Hardware
Read a stationary IMU log: bias, scatter and Allan deviation
Compute mean, standard deviation and non-overlapping Allan deviation without overinterpreting a short gyro log.
Run in your browser
Explore conditions & results →
Sensors & Hardware
Sampling and aliasing: why faster polling is not enough
Reproduce identical samples from 12 Hz and 8 Hz signals and distinguish ODR, bandwidth, polling and decimation.
Run in your browser
Explore conditions & results →
Control Engineering Primer
Experiment with PID in Python — gains, delay and saturation
Reproduce how measurement delay, actuator limits and integral windup change PID tracking on the same second-order model.
Run in your browser
Explore conditions & results →
Notes
Home-server backup and recovery — a small restore exercise
Distinguish restored files from usable application data with a small SQLite exercise, then map the checks to a real server.
Run in your browser
Explore conditions & results →
Sensors & Hardware
Robot power design — diagnosing computer resets during motor startup
Relate motor load changes to voltage drops and locate faults across the battery, wiring, converter and computer.
Run in your browser
Explore conditions & results →
Robotics Primer
From mapping to navigation in ROS 2 — a minimal Jazzy and Nav2 procedure
Use the Jazzy simulation to check mapping, map saving, reloading and goal navigation as separate steps.
Run in your browser
Explore conditions & results →
Robotics Primer
Why sensor fusion fails — timing, frames and extrinsic calibration
Diagnose fusion failures through timestamps, coordinate frames, mounting geometry and covariance.
Run in your browser
Explore conditions & results →
Robotics Primer
How to evaluate SLAM — ATE, RPE, runtime and failures
Use synthetic trajectories to distinguish absolute and relative errors, then define matching, alignment and failure reporting for an evaluation.
Run in your browser
Explore conditions & results →
Sensors & Hardware
How GNSS Works and Key Products — u-blox, Trimble, Fixposition
RTK combines satellite positioning with correction data from a base station, shrinking the several-meter error of standalone GNSS positioning down to centimeter level. We work through how dual-frequency positioning cancels ionospheric delay, how u-blox's ZED-F9P drove low-cost RTK into the mainstream, Trimble RTX and Japan's own Michibiki (QZSS), and the real-world example of Kubota's nationwide network of 342 RTK base stations for farm machinery — and how each of these changes what positioning accuracy you actually get.
Run in your browser
Explore conditions & results →