Not textbook demos. Experiments that hit real constraints.
Paths a car can actually drive, MPC that knows its steering lag, light that bends differently at every wavelength. Change the conditions and find where each method breaks.
All experiments
MPC Path Tracking
See how predictive steering dynamics smooth high-speed path tracking.
Control Lab
Pit two sampled PIDs against each other on a motor, car, cart-pole and quadrotor, with Bode, Nyquist, poles, relay tuning and an IAE gain map.
Laser Lab
A polarization ray tracer with Sellmeier glasses and Fresnel optics: measure minimum deviation, achromat vs singlet focus, and Brewster’s angle.
Robot Arm Lab
Drive four identical arms with PID, gravity compensation, inverse-dynamics feedforward and computed torque, and compare tracking under unknown payloads, torque limits and singularities.
Cosmic Forge
N-body gravity workbench: measure energy conservation per integrator and test real orbits against Kepler.
Wind Tunnel
Measure drag, lift and Strouhal number on cylinders and NACA airfoils in a D2Q9 TRT lattice Boltzmann tunnel, with Re, grid and angle-of-attack sweeps against published data.
Wave Lab
Solve 2D acoustic FDTD for double slits, phased-array steering and grating lobes, Doppler and Mach cones, and room modes, and check each against theory.
Sensor Lab
Push an image through a diffraction-limited lens, OLPF, Bayer CFA, photon noise and three demosaics, measure it with a slanted-edge MTF and a photon-transfer curve, and find where diffraction meets aliasing.
Parking Planner
Hybrid A* with Reeds-Shepp shots plans parallel, bay and K-turn manoeuvres, then a tracking controller drives the smoothed path into the slot.