Move people without breaking place
Model ferry pulses, event days, staff needs, bike and buggy access, walking links and quiet times without pushing more pressure onto fragile places.
Let The Island Rehearse Before The Surge
The old capsule hotel deck imagined city-simulating computers beside short-stay accommodation. The Straddie version is sharper: a place where locals, visitors, workers and agents can practise the next hard day before it arrives, powered first by idle capsule GPUs and eventually by a serious local compute rack.
Turn waste into a design problem
Use the lab for waste routing, food rescue, shared laundry, supply-chain visibility, maintenance planning and public works scenarios.
Practise health pressure while calm
Test temporary staff beds, intake forms, public updates, privacy rules, telehealth visits, wellness education and emergency rest capacity.
Keep disaster kiosks rehearsed
Use local compute to pre-build maps, emergency bulletins, LoRa message packs, offline notices and kiosk training scenarios before the weather turns.
Train locals on the tools reshaping work
Help residents, workers and small businesses use AI for planning, records, grant drafts, public notices, service design and local media.
Make every rehearsal leave evidence
Every scenario should leave a source-dated Markdown file that a human, agent team, council reviewer or trusted reader can reopen later.
A simulator should change behaviour, not decorate a wall.
The point is not a glowing dashboard. The point is better decisions: where capacity is thin, what a surge would break, which services are duplicated, what visitors actually need and which small investment saves the most trouble.
Inputs
Bookings, arrivals, events, weather, public notices, feedback, service capacity, source dates and local observations.
Compute
Idle capsule GPUs, kiosk nodes, noticeboard devices and future rack jobs queued by public value, privacy and urgency.
Outputs
Scenario notes, risk maps, staffing assumptions, access blocks, kiosk packets, grant appendices and public-safe dashboards.
Guardrail
No raw health, family, identity or commercial data in public dashboards. Use consented summaries and group thresholds.
From Civic Twin To Web3 Simulacrum
The compute story is not only local convenience. It is a ladder from practical island rehearsals to a larger sovereign simulation commons.
Capsule GPU pool
Small jobs: notice rendering, local LLM summaries, image/video drafts, route maps, grant appendices and scenario batches.
Disaster kiosk mesh
Hardened nodes: cached data, survival messages, emergency maps, sensor fusion and local coordination when the cloud is fragile.
Virtual Minjerribah
Island twin: ferry pulses, events, waste, food, microgrid pressure, environmental readings, heritage layers and policy sandboxes.
Web3 Sensorium
Global commons: signed data, CRDT sync, decentralised identity, open debate, science simulations and model provenance.
Supercompute rack
Heavy local jobs: custom language models, image/video/song models, scientific workloads and high-fidelity simulation once demand is proven.
Aura guardrail
The more powerful the compute, the stronger the consent shell must be: private data stays private, public results stay aggregated.