Case study

Cold Engine

Completed playable Unreal Engine demo about running a corporate train through a frozen world of debt, quotas and scarce resources.

Playable Steam demo 5-month production cycle Small team UE5 / Blender / Figma Optimization challenge
Hero gameplay capture30-60 sec gameplay or the Cold Engine trailer.
StatusCompleted playable demo / public-ready release
PeriodJanuary 2026 - April 2026, 5 month cycle
TeamSmall team: lead developer, core programmer, composer
My roleLead developer, game designer, 3D artist, technical artist
ToolsUnreal Engine 5, Blender, Figma, DaVinci Resolve
Proof linksPlay demo, gameplay capture, UI screens, optimization breakdown

Overview

A small experiment that became a finished demo.

Cold Engine started as a short capability check and grew into a complete demo. The project tested whether a small team could build, present and release a playable first-person survival-management experience in Unreal Engine.

The player manages a train in an industrial frozen world, balancing deliveries, energy, temperature, cargo pressure and corporate quotas while moving between stations.

Gameplay context boardWorld, train, delivery pressure and resource constraints in one visual.

Production story

From one-week capability test to a public playable demo.

The roadmap keeps the process visual: each step has a concrete media frame and a clear production decision.

01 / Prototype

Find the playable promise first.

The project began as a compact test around train pressure, delivery goals and first-person resource management.

Prototype captureEarly layout, train cabin and delivery readability.
02 / Route loop

Make the contract loop readable.

Contract choice, route danger and delivery stakes were shaped into a short loop a player can understand quickly.

Route choice loopContract selection, route risk and delivery pressure.
03 / Cabin systems

Turn pressure into moment-to-moment play.

The cabin, cargo and train state became the player-facing layer for energy, resources and cold-world survival.

Cabin systems passEnergy, cargo and train state readability.
04 / Optimization

Stop the main technical bottleneck.

The movement setup was reworked so the team could finish the demo instead of repeatedly fighting scene cost.

Optimization diagramOld architecture, blocker, rework and result.
05 / Visual and UI pass

Shape the demo for outside viewers.

Materials, props, snow mood, CRT-style interface pieces and video capture were brought together for public presentation.

Visual presentation frameLighting, UI, atmosphere and train readability.
06 / Public output

Finish the slice as something people can inspect.

The work ended as a complete Steam-facing demo with systems, visuals, audio, presentation material and a clear role split.

Final demo captureTrailer, Play demo page, final gameplay or release proof.

Main challenge

Optimization became the main production problem.

Problem

The original setup kept the train and player static while the world simulated movement, making the scene heavy and hard to scale.

Decision

The movement approach was reworked so the team could stop fighting the same performance blocker.

Result

The main bottleneck was removed, making the demo stable enough to finish and present publicly.

Optimization before-after diagramOld movement architecture, production problem, new solution and result.

Project takeaways

The visible result, the decisions behind it and the parts I directly owned.

A quieter summary of what the project says through the work itself.

Finished Output

A playable demo was scoped, assembled, presented and pushed to public-facing release material.

Production Decisions

The project stayed centered on one readable loop instead of expanding beyond what the team could finish.

Technical Judgment

The optimization issue was treated as an architecture problem, not just a polish task.

Collaboration Boundaries

The page separates my design, art and production ownership from programming and audio collaboration.

Next step

Want the short version?

I can walk through the build, media, role split and production decisions in a focused conversation.