StpVelox (BotUI)

The Flutter desktop environment for the KIPR Wombat — dashboard, programs, sensors, and more.

📖 Full documentation at raccoon-docs.pages.dev

---

StpVelox is the graphical environment that runs directly on the KIPR Wombat controller. It replaces the default Wombat UI with a fast, touch-friendly Flutter interface purpose-built for Botball competition — real-time sensor graphs, a program runner with live terminal output, WiFi management, and a camera feed, all on an 800×480 display.

---

Features

Screen	What it does
Dashboard	Live sensor overview grid — IMU, analog/digital ports, battery
Programs	Browse, launch, and monitor robot programs with a live terminal
Sensors	Real-time charts for gyro, accelerometer, magnetometer, and analog inputs
Camera	Live camera feed from the robot's vision system
WiFi	Scan, connect, and manage networks via Linux NetworkManager
Settings	Device config, touch calibration for the embedded display

---

Building & deploying

StpVelox uses FVM for Flutter version pinning and flutterpi_tool for ARM64 cross-compilation.

Prerequisites

# Install FVM
dart pub global activate fvm
fvm install   # picks up version from .fvmrc

# Activate flutterpi_tool
dart pub global activate flutterpi_tool

Build & deploy to Pi

# Build release + deploy to Pi in one step
DEPLOY_HOST=<your-pi-ip> bash deploy.sh

Debug (profile mode on connected Pi)

bash debug.sh

Build only (output in build/)

bash build.sh

The build script runs build_runner for code generation (Riverpod, Freezed, JSON serialization), then cross-compiles for arm64 / pi3.

---

Architecture

StpVelox follows Clean Architecture with three layers and Riverpod for state management throughout.

lib/
├── core/           # DI (GetIt), routing (go_router), theme, shared utilities
├── features/       # One folder per feature (dashboard, program, sensors, wifi, ...)
│   └── <feature>/
│       ├── domain/       # Entities, repository interfaces, use cases
│       ├── data/         # Repository implementations, data sources
│       └── presentation/ # Riverpod providers, screens, widgets
├── shared/         # Shared widgets and helpers used across features
└── main.dart

Hardware data flows in via LCM through raccoon-transport — sensor values, motor state, and IMU readings are published on the LCM bus and consumed by the UI in real time.

Program execution uses a pseudoterminal so program stdout/stderr stream live into the terminal widget (xterm).

Rendering uses flutter-pi to render directly to the Linux framebuffer — no desktop environment, no compositor overhead. This gives high frame rates at almost no system cost on the Pi.

---

Requirements

• FVM — Flutter version manager
• Dart SDK ^3.5.4
• flutterpi_tool for ARM64 builds
• KIPR Wombat with Raspberry Pi (ARM64, 800×480 display)

---

Part of RaccoonOS

Repository	What it is
raccoon-lib	Core robotics library
raccoon-cli	Dev toolchain — scaffolding, raccoon run
raccoon-transport	LCM messaging layer
documentation	Full platform docs

---

License

Copyright (C) 2026 Tobias Madlberger
Licensed under the GNU General Public License v3.0 — see COPYING for details.