Experimental & Interactive Media

01Build a webcam body/hand-tracking interaction in the browser with MediaPipe or ml5.js where a gesture drives a visible response (hand position scales/colors a shape).

02Define a gesture vocabulary for an imagined installation: 3 gestures, each with an ergonomic note (can a tired or short person do it?), a dead-zone, and a 'rest' state that doesn't trigger by accident.

03Add smoothing/filtering to your activity-1 tracker (moving average or one-euro filter) and document the jitter reduction vs added latency tradeoff.

01Drive an addressable LED strip (NeoPixel/WS2812) from an Arduino/ESP32 with 3 lighting 'looks' and one audio- or motion-reactive mode.

02Patch a virtual DMX fixture in QLC+ and program a cue list (chase + a manual override), running it against the built-in visualizer with no hardware.

03Write an interactive-lighting design + safety checklist: power budget for your strip length, heat, flicker/photosensitivity caution, and a fail-safe default state.

01Wire one sensor (PIR motion, ultrasonic distance, or a capacitive touch pad) to an Arduino, print clean readings to serial, and trigger an output when a threshold is crossed.

02Send your sensor data over OSC into Processing/TouchDesigner and map it to a visual parameter, confirming end-to-end latency feels acceptable.

03Characterize and treat your sensor's noise: log raw data, apply a threshold + debounce + smoothing, and document false-trigger rate before/after. For depth, note that Azure Kinect/RealSense are discontinued and plan around Orbbec Femto Bolt or webcam tracking.

01Compose a 3-5 minute generative/ambient soundscape in Pure Data (or a free DAW) with at least two layers that evolve independently so it never loops audibly.

02Design a multi-speaker spatial layout for a real room: plan source placement, listener path, and which sounds a person hears as they walk a route; simulate with 2 speakers + panning if you can't build it.

03Make the soundscape reactive: feed one sensor/OSC value (presence or distance) into Pd to shift the mix when a 'visitor' approaches.

01Write a public-installation production plan for a real site: power source, mounting, weatherproofing (IP rating), sightlines, accessibility, daily maintenance, and an unattended fail-safe.

02Storyboard the unsupervised interaction: how a passer-by understands it in 5 seconds, what stops vandalism/misuse, and how it resets between users.

03Build a small weather/operations runbook: a startup/shutdown sequence, a 'what to check each morning' list, and a remote heartbeat/monitoring idea.

01Build a tiny WebXR VR scene (A-Frame) the viewer can look around in, with one grabbable/gaze-triggered object and a comfortable default scale.

02Recreate a Unity XR Interaction Toolkit starter scene against OpenXR and add one comfort feature (teleport locomotion or vignette on turn).

03Write a VR comfort + presence checklist: frame-rate floor, locomotion options, UI distance, snap-turn, and a seated fallback; map each to the discomfort it prevents.

01Sew a simple soft circuit: a coin-cell, conductive thread, and one LED on fabric that lights when a sewn switch/snap closes.

02Recreate an Adafruit wearable guide (e.g. sound- or motion-reactive NeoPixel accessory) on a sewable microcontroller and adapt the animation.

03Document a wearability + safety pass: strain relief at flex points, battery safety, washability/removal plan, and skin comfort; redesign one weak point from activity 2.

01Build a phone-as-controller crowd interaction: a web page audience members open to vote/move/color a shared big-screen visual in real time via WebSockets.

02Write a crowd-dynamics + load production note: what happens at 5 vs 500 users, rate-limiting, dominant-user prevention, and a graceful-degradation plan.

01Capture one live biosignal (heart rate via a cheap pulse sensor, or breath via a stretch/mic proxy) into Arduino, then drive a visual whose intensity tracks the signal.

02Draft a participant consent + data-handling protocol for biometric art: what is captured, that nothing is stored, opt-out mid-experience, and a per-person calibration step.

01Drive a vibration motor / LRA from a microcontroller and design 3 distinct haptic 'words' (a tap, a rising alert, a heartbeat) tied to digital events; test if a blindfolded person can tell them apart.

02Measure and tune event-to-haptic latency: log the delay between the digital trigger and the felt buzz, then minimize it and document the perceptual threshold you targeted.

01Build a cue-driven A/V rig where a single trigger (MIDI/OSC/keypress) advances synchronized visuals + audio + lighting, rehearsed against a fixed timeline.

02Write a show production case study: signal-flow diagram, a failure-recovery plan (what the operator does if a cue misfires live), and a soundcheck/tech-rehearsal checklist.

01Build a custom tangible controller from cheap inputs (a potentiometer, a force sensor, a rotary encoder) that maps physically to a digital parameter people can read without instructions.

02Iterate the controller's industrial-design feel: prototype 2 enclosure/knob variants, test grip and labeling with a user, and document which felt right and why.

01Build a two-location synced visual: send OSC/WebSocket data between two machines (or two browser tabs simulating remote sites) so an action in A appears in B, and measure the round-trip latency.

02Write a latency-as-material design note: choose to hide, embrace, or compensate for delay, and define a sync strategy (shared clock vs event-driven) with a degraded-network fallback.

01Design and prototype a small multisensory 'pod': one enclosed corner where synchronized sound, light, and one tactile or scent cue respond to a participant's presence, all co-designed to a single emotional beat.

02Produce an integration case study analyzing where your pod's channels almost desynced and how you bound them to one clock/trigger, plus the accessibility accommodation for a participant who can't use one sense.

01Build a complete small interactive installation: a sensor input drives synchronized visuals and sound in a defined space, documented end-to-end as a working system (approach -> response -> reset).

02Write the install/reinstall + maintenance documentation pack for that installation so a different technician could set it up without you: bill of materials, wiring, calibration, startup/shutdown, and teardown.

01Stage a one-performer mixed-reality scene: body/marker tracking (webcam/MediaPipe) drives real-time projected or screen visuals that respond to the performer's movement, rehearsed to a fixed cue sequence.

01Build a small kinetic piece: a microcontroller drives a servo/stepper to choreograph a repeating motion that changes when a sensor detects a person, with a defined safe range of motion.

01Build a free Pepper's-ghost / reflective pyramid volumetric illusion: model a rotating 3D form in Blender, render the 4-view layout, and display it through a clear acrylic/CD-case pyramid on a phone or monitor.

02Write a practice essay/spec proposing a volumetric piece for real hardware (e.g. a Looking Glass-class light-field display or a swept-volume display), naming the technical constraints and how your content design works with them.