GasFlow — Mobile Delivery Operations

Mobile logistics system built with React Native, Rust, Axum and PostgreSQL for scheduled gas-cylinder deliveries, stock control and operational traceability.

GasFlow digitalizes a local distribution workflow where orders are scheduled in advance, assigned to drivers and reconciled against full and empty cylinder movements. The product prioritizes operational control over instant-delivery marketplace behaviour.

Status: functional MVP under active development. Core backend and mobile workflows are implemented; verified screenshots, a distributable mobile build and a hosted demonstration backend remain pending.

Business problem

Small delivery operations often coordinate orders, routes, stock and returns using calls, messages and manual notes. That creates avoidable problems:

• Orders are difficult to prioritize and assign.
• Delivery status is not consistently recorded.
• Full and empty cylinders lose traceability.
• Failed deliveries and rescheduling are handled informally.
• Daily operational reporting requires manual reconciliation.

GasFlow converts those activities into explicit workflows and verifiable records.

What the MVP supports

Administration

• Schedule orders by date and delivery window.
• Filter and paginate orders.
• Assign orders to drivers.
• Register inbound full-cylinder stock.
• Review stock summaries and daily reports.

Driver operation

• Authenticate and view assigned orders.
• Record successful deliveries.
• Register full cylinders delivered and empty cylinders received.
• Record failed deliveries with an optional reschedule.

Operational control

• Order-status transitions.
• Request IDs for HTTP traceability.
• Audit events for critical changes.
• Health and metrics endpoints.
• Network-aware mobile state and local persistence support.

What this repository demonstrates

Area	Evidence
Mobile development	React Native, Expo, TypeScript and role-oriented navigation
Backend engineering	Rust, Axum, Tokio, SQLx and PostgreSQL
Architecture	Hexagonal modular monolith with domain, application, ports and adapters
State and data access	TanStack Query, Zustand, AsyncStorage and NetInfo
Security	JWT authentication and bcrypt password hashes
QA	Rust tests, mobile tests, type checking and GitHub Actions
Operations	Docker Compose, migrations, metrics, health checks and request tracing

Architecture

flowchart LR
    ADMIN[Administrator] --> APP[React Native app]
    DRIVER[Driver] --> APP
    APP -->|REST / JSON| API[Rust + Axum API]

    subgraph Backend
        API --> APPLICATION[Application use cases]
        APPLICATION --> DOMAIN[Domain rules and ports]
        APPLICATION --> ADAPTERS[HTTP, auth, DB and observability adapters]
        ADAPTERS --> DB[(PostgreSQL + SQLx)]
        API --> AUDIT[(Audit events)]
        API --> METRICS[Metrics and request IDs]
    end

Loading

The backend is deployed as one modular service. Hexagonal boundaries isolate business rules from Axum, JWT and PostgreSQL adapters, preserving a practical path for future evolution without introducing premature microservices.

Technology stack

Mobile

• React Native 0.76 and Expo 52.
• TypeScript.
• React Navigation.
• TanStack Query.
• Zustand.
• AsyncStorage and NetInfo.
• React Native Testing Library and Jest.

Backend

• Rust and Axum.
• Tokio async runtime.
• SQLx and PostgreSQL.
• JWT and bcrypt.
• Utoipa / Swagger UI.
• Structured tracing, metrics and request IDs.

Infrastructure and quality

• Docker and Docker Compose.
• Versioned database migrations.
• GitHub Actions.
• Backend tests against PostgreSQL.
• Mobile type checking and automated tests.

Main API flows

Authentication

• POST /auth/login
• GET /me

Orders and dispatch

• POST /orders
• GET /orders?date=&status=&assignee=&page=&page_size=
• PATCH /orders/{id}/status
• POST /dispatch/assign

Delivery execution

• POST /deliveries
• POST /deliveries/failed

Stock and reporting

• POST /stock/inbounds
• GET /stock/summary?date=
• GET /reports/daily?date=

Operations

• GET /health
• GET /metrics
• Swagger UI at /swagger-ui

Quick start

Backend and database

git clone https://github.com/sjo1848/gasflow.git
cd gasflow
docker compose up -d --build
docker compose ps

Services:

Service	Address
Backend	http://localhost:8080
PostgreSQL host port	localhost:5433

Stop the stack:

docker compose down

Mobile application

cd mobile
npm install
EXPO_PUBLIC_API_BASE_URL=http://localhost:8080 npm run start

For an Android emulator, the API host may need to use the emulator-specific host address rather than localhost.

Validation

Backend

cd backend
cargo fmt --check
cargo test

Mobile

cd mobile
npx tsc --noEmit
npm test -- --runInBand

The CI workflow runs backend tests against a PostgreSQL service and validates mobile type safety and tests on every push and pull request to main.

Development credentials

The development seed includes example administrator and driver accounts. These credentials are for local development only and must never be reused in a public or production deployment.

Documentation

• Portfolio case study
• Implementation status
• Product requirements
• Architecture
• Backlog
• Architecture decisions

Current priorities

• Add verified mobile screenshots and a short walkthrough.
• Produce a reproducible Android build.
• Validate intermittent-connectivity behaviour through automated scenarios.
• Add a controlled demonstration backend.
• Test workflows with realistic delivery volumes and stock reconciliation cases.

Scope note

GasFlow is a portfolio-grade MVP for a specific local distribution workflow. Production use would require deployment-specific security, privacy, monitoring, support, device management and operational validation.