FMX 430 · Spatial Computing

How this course works

The course opens with the whole earth in the viewport and a very old problem. In Borges' On Exactitude in Science, cartographers build a map at 1:1 with the empire it describes — and the map, being useless, is left to rot in the desert. We now build those maps for real. Your first project is one.

You will build things with your own hands before you direct anything to build them for you. Weeks 1–8 are hands-on only — Unreal, Twinmotion and the geo stack, no AI agents in the work. From Week 9 the CLI agents join the studio and stay to the end of term.

That order is deliberate. The hands-on is the learning, and you cannot judge whether an agent did good work until you can do the work. The AI TA on the course site stays available the whole term for questions — asking why is my Lumen scene noisy is help; having an agent build the level is not.

Grade breakdown

Scale. 95–100 A · 89–94 A/B · 84–88 B · 77–83 B/C · 72–76 C · 66–71 C/D · 60–65 D · below 60 F.

The semester

Fourteen Friday studio sessions plus a finals-week exhibition, in four phases. Every week links to its own page — the brief, the guided lab, the homework and the resources live there.

No class Nov 27 (Thanksgiving).

The Friday rhythm

We meet once a week for four hours. Every session is the same four blocks — only the third one changes at Week 9.

Projects

P1 · Digital Twin Flyover

10% · Weeks 1–3 · Due W3 crit · hands-on, no agents

You begin with the whole earth. Set up Cesium for Unreal, pull real terrain and photogrammetric building data at true coordinates, choose a landscape anywhere on the planet, and make a 90-second cinematic flyover of it.

The camera is the assignment as much as the terrain. Sequence a Cine Camera Actor so the move is legible and unhurried — no sudden turns, no rapid movement. A flyover that lurches tells the viewer nothing about the place.

Deliverables. A georeferenced Unreal level with your Cesium tilesets · a 90-second rendered flyover · the coordinates of your site documented in your project folder · a visible data attribution for every source you streamed · your process log.

P2 · Architectural Visualization

15% · Weeks 3–6 · Due W6 crit · hands-on, no agents

Having flown over the real world, you now build inside it. A designed space at strict 1:1 — measured, blocked out, brought through Twinmotion into Unreal via Datasmith, materialled honestly and lit to a specific hour.

Sited, not floating. The building stands at true coordinates on Cesium World Terrain with real neighbors and a matched sun — a before/after pair and one 30-second approach shot in context. Siting quality is a rubric row.

Deliverables. A navigable level at true scale · an interior still set (5 or more) · two honest lighting states · the sited context shots · a 60-second walkthrough · process log.

P3 · Guided training practical exam

20% · Week 7 · Oct 16 · No AI of any kind

Stations, timed, hands-only: no agents, no chat, no AI TA, no internet. The station set includes one geo station — georeference, anchor to the globe, set the sun to the exam hour, screenshot — using Cesium World Terrain and OSM Buildings only, so nothing depends on a Google quota under time pressure.

Weeks 1–8 are hands-on precisely so this exam tests what you have actually been doing.

P4 · Experimental Project

15% · Weeks 8–12 · Due W12 crit · agents allowed from Week 9

The open one. An immersive or experimental work that is yours — a piece of spatial self-expression rather than a specification met. It can be a VR piece, a projected or LED-wall installation, a data-driven landscape, a VP Studio collaboration, or something the brief did not anticipate.

It launches the week before the CLI opens, deliberately: you scope it by hand, then build it with agents available from Week 9 and log what they did.

Deliverables. A working build or installation · a statement of intent (300 words) written at launch, before you build · documentation that shows the piece as experienced · process log, plus the AI log from Week 9 on.

P5 · Final Project

40% · Weeks 12–15 · Exhibited finals week, VP Studio

Your voice as a maker of spatial work, at the scale the VP Studio allows. The Final is a choice, and any thread from earlier in the term is an approved direction to scale up — the Experimental Project most obviously, but also a site you flew over in P1 becoming a place someone can stand in.

Build an immersive world. The default, and the one most students take: a place with intent, made to be entered rather than watched. Navigable on screen, on the LED wall, or in a headset.

Or build for a headset. HTC Vive and Meta Quest units are available for student use — you do not buy anything. See the fallback rule under P4: it applies here too, and it is not negotiable in an exhibition week.

Or make a snowglobe. A seasonal option, and a better design brief than it sounds: design a snowglobe and let the player navigate the inside of the sphere. Screen-playable, no headset required. The bounded spherical world is a real constraint — there is no horizon to hide behind, every direction is a designed view, the interior camera has to behave inside a curved shell, and scale reads strangely by definition. Small, finishable, and genuinely hard to do well.

Deliverables. A working build or installation · a statement of intent · documentation that shows the piece as experienced · the full process log, with the AI development log covering Weeks 9–15.

Exhibition formats. LED wall (cinematic and interactive, architectural scale) · gamepad stations (navigable builds) · ambient projection · at most one attended VR station, if earned.

The Week 15 pre-show audit is mandatory: attribution visible on every piece that streamed third-party tiles, data-source credits on data pieces, and Cesium ion tokens valid through the run.

Course AI policy

Weeks 1–8 — hands only. No AI agents in the work: no Claude Code, no Codex, no generative tools producing geometry, materials, levels or code you submit. Ask the AI TA anything you like about how something works; do the work yourself.

Week 7 — nothing at all. No agents, no chat, no AI TA, no internet. The practical exam is unaided by design.

Weeks 9–15 — allowed and expected, and logged. Use Claude Code and Codex on production work, with every session recorded in your AI development log: prompt, what the agent did, what you kept, fixed or rejected.

Required all term. You must be able to explain every asset, material, level and line of code in your submitted build — live, at crit. "The AI did it" is a starting point, not an answer.

Never allowed. Presenting agent output as hand-authored · using AI to fabricate references, data sources or reflection writing · using agents on Week 1–8 submitted work or during the Week 7 exam.

The geo track is the best verification gym in the course, because the ground truth is a real place: when the agent is wrong about the sun over Tampa, you can go outside and check.

Attendance, crit and conduct

Attendance. This is a studio course: the work happens in the room, and every session ends in a crit. More than three unexcused absences results in a grade of NF.

Excused absences are different and are accommodated, per the University's attendance policy. Tell me in advance for anything scheduled — athletics, religious observance, University business — and as soon as you can for anything unscheduled. Missed work will be replaced with something equivalent or excluded from your grade. Documentation may be required. Note the University's own caveat: if too much participation-dependent work is missed, learning outcomes may not be achievable even when the absences are excused.

Crit conduct. Crit is the engagement expectation in this course. Screens up, work shown as it is rather than as you wish it were, and feedback that is specific and about the work. Describe what you see before you say what you would change.

The studio and the VP Studio. Everyone here has a right to a constructive learning environment. The Cass Building VP Studio has its own access and safety rules, which are not optional. Behavior that disrupts the class is addressed under the University's Disruption of the Academic Process policy.

Equipment & software

• No laptop required. Every student has a lab PC in CCB-134 that runs Unreal Engine 5.8, available for all class work. Bring your own laptop if you prefer, or work on a gaming desktop at home between classes — for personal machines a discrete GPU is strongly recommended.
• Free account: Epic Games (UE 5.8 and Twinmotion).
• A Cesium ion account — needed in Week 1. Keep your token valid through finals week.
• Claude Code and Codex CLI — installed in Week 9, not Week 1.
• Lab headsets are provided — HTC Vive and Meta Quest units are available for student use. VR is an option inside P4 and the Final, never a requirement, and you are not expected to buy anything.
• Suggested: a three-button mouse, a 1 TB external SSD, and a USB-C data cable for headset work.

University policies

These are University of Tampa policies. They apply whether or not they are reproduced here, and they are stated in full on the official syllabus in SpartanLearn (Canvas), which governs.

Who to contact