Barnicle Cleaning Game Roadmap

01 · Where it stands

A strong mechanic and one missing loop.

The prototype already proves the hard part: a hull that can be cleaned, barnacles that can be removed, and tools that feel different. What it does not have yet is a reason to finish, a score to read, or a way to keep progress after quitting.

3Working tools

1Cleanable boat

3Barnacle volumes

0Complete jobs

Works today

One hull with three barnacle volumes and physical growth in each.
Default scraper, wire scraper and hydro jet with real, different footprints.
Money, buying and equipping tools through the tablet shop.
A wet-sand shoreline, ocean water and PCG beach debris.

Half-wired

Contracts screen exists as UI; no live job data or payout behind it.
Per-volume barnacle counts are tunable, but density and scale seeds are not wired.
Brush direction, shape and removal already match the painted footprint.
The beach looks good; foam and shoreline effects are parked for later.

Missing entirely

Job completion, ratings and payouts.
Save and load, including the painted state of a hull.
Scrape audio, chips, spray and a completion moment.
A second boat, a tutorial, settings and a packaged build.

02 · The game I would build

A cozy boatyard, run one hull at a time.

You inherit a run-down boatyard on a small island. Boats arrive filthy and leave gleaming. You read the work order, pick the right tool, and work the hull by hand until the owner can see their reflection in it. The tide sets the schedule. The yard grows as you do. That is the game: a physical job sim with no combat, no fail screens, and a craft you get visibly better at.

The tool in your hands

Resistance, angle, sound, flying chips, spray. The scrape is the game, so it has to feel better than it looks.

The reveal

Every job is built around the moment the hull comes back. Before and after is the reward, so protect that moment.

The yard that grows

Money becomes tools. Reputation becomes contracts. Earnings become a marina you are proud to show.

03 · The loop

Three cycles, stacked.

Every feature in this plan exists to support one of these three loops. If a feature does not make one of them better, it belongs on the cut list.

30 seconds

The patch

Find the next dirty patch
Match the tool to the grime
Work it from the right angle
Watch it come off

Repeat until the zone is clean.

10 to 20 minutes

The job

Read the contract
Inspect the boat and set up
Clean every zone to threshold
Do the finishing details
Submit for inspection

One session

The season

Bank the pay
Buy or upgrade a tool
Take a bigger boat
Grow the yard

Test the loop with grey boxes before adding art. If cleaning one ugly cube is not satisfying, ten beautiful boats will not fix it.

04 · Everything the game should have

The systems catalog.

Grouped by the part of the game it serves. Tags say when it belongs: Now blocks the first playable loop, Next belongs in the vertical slice, Later is full-game scope, and Maybe only earns a place if the core is already fun.

Nowfirst playable loop Nextvertical slice Laterfull game Maybeonly if the core sings

Build note Have one brush stamp update two things: the visual grime mask and a small cleanliness grid per zone. The grid is the score, the mask is the picture. Grid cells are cheap to save, easy to compare against a threshold, and they avoid reading pixels back from a render target.

System A

Cleaning and tools

The core verb. Everything else exists to give it context.

Zone cleanliness tracking. Split each hull into named zones such as bow, mid, stern, waterline and bottom. Each zone reports a clean percentage, and the job system reads those numbers.Now
Grime types. Salt film, algae, rust, oil, old antifouling and mussel clusters. Each one prefers a different tool, so the player makes a real choice.Next
Tool feel and control. Angle, pressure, distance and dwell time. The jet strips soft grime, the blade pops barnacles, the wire brush cuts rust but scars paint.Next
Damage and care. The wrong tool on the wrong surface leaves scratches and gouges. Warn the player clearly, then charge for the repair at inspection.Next
Stubborn growth. Big clusters need several passes from the right angle. Rust needs chemical dwell, then a rinse. Not every spot should fall to one swipe.Next
Clean-vision. A toggle that softly highlights remaining grime. It prevents pixel-hunting and makes the game kinder to play.Next
Tool tiers. Three tiers per tool family. Each tier should change the feel and reach, not just the damage number.Later
Surface materials. Fiberglass, teak, steel and aluminum have different hardness, damage rules and ideal tools.Later
Finishing services. Polish, wax and antifouling paint as paid add-on services. They raise the rating and give late-game jobs a finishing act.Later

System B

Boats and dock

The objects you clean and the places you work on them.

Reusable cleanable hull. Turn the current HullPainter into one component: volumes, grime mask, barnacle scatter and per-boat state. Any hull can then be made cleanable.Now
The first three boats. A small skiff, a fishing boat and a sailboat. Different hull shapes, sizes and access, so jobs start to feel different.Next
Dry dock lift. Boats arrive in the water or on a trailer. The lift exposes the bottom, which gives the waterline and the tide a real purpose.Next
Boat presentation. Gentle bobbing at the dock, mooring lines, fenders and a correct waterline mask. Small details that sell the boat as a real object.Next
Decks and interiors. Teak decks, cabins, windows and cushions. A second surface type that supports a full-detail contract.Later
Repair jobs. Zincs, prop dings and cracked planks. Inspect the damage, quote the repair, then do the work.Later
Restoration arcs. The same boat returns for hull, deck and paint over several jobs. The owner remembers you, and the finale means more.Later
Drivable boats. Delivery trips and sea trials. It is fun, but it is a second game's worth of systems. Keep it out of v1.Maybe

System D

Money and progression

Why you take the next job.

Save game. Money, owned and equipped tools, completed jobs, boat dirt state and yard upgrades. Without this, every session starts from zero.Now
Payouts and costs. Show base pay, tips and bonuses on one screen, and show what consumables and repairs cost. The player should understand the margin.Next
Reputation. Stars and levels unlock better contracts, licenses and new areas. Reputation is the long-term reason to do a good job.Next
Shop expansion. Sections for tools, upgrades and consumables. Keep the card style that already works, and add an upgrade row per tool.Next
Yard upgrades. A bigger dock, a second lift, a paint tent and a better shop front. The yard is the visible save file.Later
Helper. Hire someone to collect beach debris or pre-soak hulls. Automation should feel like a reward, not a shortcut.Later
Cosmetics. Gloves, tool skins, yard decorations and boat names. A harmless money sink for players who want one.Maybe
Operating costs. Rent, electricity, staff. It adds pressure, but it can make a cozy game anxious. Test it before committing.Maybe

System G

UI, onboarding, settings

Players should never wonder what to do next.

Job HUD. Active contract, zone checklist, cleanliness percentage, pay and tide window. Readable at a glance from any camera angle.Now
Tool wheel. Hold a key, pick from a radial, swap instantly. The tablet is for buying, not for switching mid-scrape.Next
Tablet expansion. Jobs, shop, map, gallery and settings, all with real data and badges for new work.Next
Tutorial contract. One guided job: walk, accept, equip, scrape, submit. Prompts fade after the first job.Next
Save-state UX. Autosave on job completion, a visible save indicator and a safe quit flow.Next
Pause and settings. Audio, video, controls, FOV, screen shake, hold versus toggle and text size. Basic accessibility ships with the slice.Next
Damage feedback UI. Visible scratches, a warning tone and an inspection note that explains what happened. Punishment without explanation is just confusion.Later
Localization and glyphs. No hard-coded strings in widgets and automatic controller glyph switching. Cheap now, expensive later.Later

System H

Technical foundation and shipping

Make content cheap to add and builds safe to share.

Data-driven definitions. BoatDef, JobDef, ToolDef and ClientDef as data assets. Adding a job should not require editing a blueprint graph.Now
Mask serialization. Save the painted hull mask along with the regular save data. A half-cleaned boat must still be half-clean tomorrow.Now
Smoke tests. Map loads, a job completes, money changes, save and reload keeps the mask. Run these before every build you share.Next
Performance budget. Measure early: 60 fps on a mid-range PC, barnacle HISM caps, Nanite and Lumen toggles, PCG culling.Next
Packaged builds. Ship a zipped build after the vertical slice. Playtests need a build, not a session in the editor.Next
Tuning data. Log job time, tools used and damage taken. Tune pay and cleanliness thresholds from real numbers.Later
Store prep. Steam page, capsule art, trailer, achievements and cloud saves. Later, but it has a lead time.Later
Co-op. A natural fit for this genre, and a v1 architecture trap. Keep systems actor-based so it stays possible, but do not build it yet.Maybe

05 · Roadmap

Five phases, in the order that de-risks the game.

Phase 0 is the one that matters this week. Phase 1 is the one worth showing people. Everything after that is the game you want to make if the first two prove the loop.

Rough estimates for one developer with heavy AI assistance. The order matters more than the dates, and you can stop after Phase 1 with a demo worth showing.

00start
Close the loop
1 to 2 focused weeks

One contract, from walk-up to payout, with progress that survives a restart.
- Zone cleanliness and completion event
- Reusable CleanableHull component
- JobDef plus job state machine
- Contract board wired to real jobs
- Submit, rating and payout screen
- Save game and hull mask serialization
- First scrape audio and particles
Done when A player cleans one boat, gets paid, buys a tool, quits, relaunches and still has the money and the half-cleaned second boat.
01slice
Vertical slice
2 to 4 focused weeks

A 15-to-20-minute demo that a stranger can finish without help.
- Three boats and five handcrafted jobs
- Tutorial contract
- Clean-vision and damage feedback
- Tool feel pass
- Before/after completion reveal
- Basic ambience and music
- Settings and a packaged build
Done when Three outside players finish it without a guide and ask to keep playing.
02core
Core game
4 to 8 focused weeks

A weekend of progression with a reason to take the next boat.
- 20-plus contracts across three tiers
- Six to eight boats
- Reputation levels and tool upgrades
- Beachcombing and debris recycling
- Tide cycle and day/night
- First yard upgrade
- Gallery of finished jobs
Done when Four to six hours of play, and players can name what they are saving for.
03world
World and story
4 to 8 focused weeks

The island feels like a place with people on it.
- Town strip and harbor office
- Eight recurring clients
- Multi-stage restoration arcs
- Weather and storm events
- Collectibles and secret areas
- Full music and ambience pass
- More yard upgrades
Done when Players remember client names and want to save the marina.
04ship
Polish and ship
3 to 5 focused weeks

A build you can put in front of paying players.
- Animation and VFX polish
- UI/UX and accessibility pass
- Performance and memory pass
- Bug fixing and playtest rounds
- Achievements, cloud saves, store page
- Trailer and demo build
Done when A hundred testers, no crash blockers, 60 fps on target hardware, store build approved.

06 · Next moves

What to build, in order.

These are the ten tasks that move the prototype from a mechanic to a game. The first four are the actual start of Phase 0.

If you only do one thing this week: make one contract payable end to end. No new boats, tools or island content until that works.

Make one job payable end to end. Zone percentage, completion check, rating, money. No new content until this works.
Save the result. Money, tools, job history and the hull's dirt state, including a relaunch test.
Refactor the hull painter into CleanableHull. Do it before the second boat, not after.
Wire the Contracts app to JobDefs. Accept, activate, submit, clear.
Add the first feedback pass. Barnacle crack, chips, jet loop and a completion chime.
Add clean-vision and a damage warning. Playability before realism.
Build boat number two and jobs two and three. Test the 15-minute loop.
Add a tool wheel and a basic settings menu. Stop opening the tablet to swap tools.
Put a before/after card at payout. Make the reveal shareable.
Playtest with someone who has never seen it. Fix the top three confusions, then package a build.

07 · Risks and guardrails

What could go wrong, and what not to build.

The biggest danger is not a missing feature. It is building the wrong feature for six weeks.

The ten-minute problem

Cleaning can feel samey once the novelty fades. Mitigate with grime types, tool choices, sound variation and early playtests. If the loop is not fun by job three, stop adding content.

The save problem

Painted masks live in render targets, and render targets do not save themselves. Prototype mask persistence in Phase 0, before the boat count grows.

The island gravity well

A town, weather, NPCs and secrets can eat months. The cut list is part of the plan, not an insult to the ideas.

No multiplayer in v1
No open world
No drivable boats until the cleaning loop is proven
No procedural contracts before ten handmade ones are fun
No more tools before one boat feels great
No island town before the vertical slice
No energy, hunger or fatigue meters
No photoreal look that costs the frame rate
No new content when the demo needs sound and feedback

08 · Content targets

How much content is enough.

These are the numbers I would hold the vertical slice and the full game to. They are deliberately small. A tight, polished slice beats a wide, unfinished island.

Content	Vertical slice	Full game	Notes
Boats	3	10 to 12	Start with one hull mesh and vary the volumes before modeling new boats.
Contract types	3	8	Scrape, full detail and beach cleanup first.
Handmade jobs	5	25 to 30	Handcraft the first ten, then template.
Tools	4 to 5	10	Feel first, count second.
Grime types	3	7	Barnacles, algae and rust carry the slice.
Recurring clients	3	8	Portraits and short lines, not cutscenes.
Textures	8 to 10	30+	Scenario free plan: 50 CU a day, and 1K is enough.
Audio cues	15	60+	Tool loops, ambience, UI and music.
Music tracks	1	3 to 4	One calm loop can carry the slice.

09 · Playtest scorecard

The seven questions that matter.

Ask these after a first session. The answers decide whether the next work is content or the loop itself.

After one minute, can the player say what the job is?
Does the first barnacle pop make them smile?
Do they know which tool to use and why?
Can they see the hull changing as they work?
Can they explain their rating and their pay?
Do they want one more job when the first ends?
Would they show the before/after to someone?
If three or more answers are no, the next task is not more content. It is the loop.

10 · Decisions to make

Where I would come down.

None of these block Phase 0. They do change how the vertical slice feels, so decide them before Phase 1 ends.

Decision	My lean	Why
Camera	Stay third-person for now	It already works. Test first-person hands once animations exist; tactile games often win there.
Failure	No hard fails	Damage and tide reduce pay and rating, never delete progress. Keep the game cozy.
Deadlines	Untimed core jobs	Tide windows only on special contracts. Pressure should be a spice, not the meal.
Multiplayer	Not in v1	It doubles the technical surface. Keep systems actor-based and revisit after launch.
Art direction	Grounded and sunny	Readable dirt and clean surfaces matter more than photoreal water. Stylized ages better.
Build order	Vertical slice before world	A small loop that is fun beats a big island nobody can play.

The fun part is built.
Now build the game.

A strong mechanic and one missing loop.

Works today

Half-wired

Missing entirely

A cozy boatyard, run one hull at a time.

The tool in your hands

The reveal

The yard that grows

Three cycles, stacked.

The patch

The job

The season

The systems catalog.

Cleaning and tools

Boats and dock

Jobs and contracts

Money and progression

Island and world

Feedback and presentation

UI, onboarding, settings

Technical foundation and shipping

Five phases, in the order that de-risks the game.

Close the loop

Vertical slice

Core game

World and story

Polish and ship

What to build, in order.

What could go wrong, and what not to build.

The ten-minute problem

The save problem

The island gravity well

How much content is enough.

The seven questions that matter.

Where I would come down.

The fun part is built.Now build the game.

A strong mechanic and one missing loop.

Works today

Half-wired

Missing entirely

A cozy boatyard, run one hull at a time.

The tool in your hands

The reveal

The yard that grows

Three cycles, stacked.

The patch

The job

The season

The systems catalog.

Cleaning and tools

Boats and dock

Jobs and contracts

Money and progression

Island and world

Feedback and presentation

UI, onboarding, settings

Technical foundation and shipping

Five phases, in the order that de-risks the game.

Close the loop

Vertical slice

Core game

World and story

Polish and ship

What to build, in order.

What could go wrong, and what not to build.

The ten-minute problem

The save problem

The island gravity well

How much content is enough.

The seven questions that matter.

Where I would come down.

The fun part is built.
Now build the game.