robo-infra 0.3.2

Universal robotics infrastructure - control any robot from servo to rocket

robo-infra

Universal robotics infrastructure for any hardware, any project.

Overview

robo-infra is a hardware-agnostic robotics infrastructure library designed to work with any robotics project—from hobby servos to industrial robots, from quadcopters to autonomous vehicles.

Key Features

• ** Universal Hardware Abstraction** - Works with any servo, motor, sensor, or controller
• ** Simulation-First** - Everything works without hardware by default
• ** AI-Native** - Built-in integration with ai-infra for LLM-controlled robots
• ** API-Ready** - Seamless integration with svc-infra for REST/WebSocket APIs
• 🛡 Safety-First - Comprehensive limits, emergency stops, and collision detection
• ** Observable** - Full telemetry, logging, and monitoring built-in

Architecture

┌─────────────────────────────────────────────────────────────┐
│                      Your Application                        │
├─────────────────────────────────────────────────────────────┤
│  Controllers    │  Motion Planning  │  Safety Systems       │
├─────────────────────────────────────────────────────────────┤
│      Actuators (Servos, Motors)  │  Sensors (IMU, Distance) │
├─────────────────────────────────────────────────────────────┤
│                    Hardware Drivers                          │
├─────────────────────────────────────────────────────────────┤
│     Buses (I2C, SPI, UART)    │    Pins (GPIO, PWM, ADC)    │
├─────────────────────────────────────────────────────────────┤
│  Platforms: Raspberry Pi │ Arduino │ Jetson │ Simulation    │
└─────────────────────────────────────────────────────────────┘

Installation

Core Only (Minimal)

For embedded systems or when you just need robotics abstractions:

pip install robo-infra

With AI Integration

For LLM-controlled robots using ai-infra:

pip install robo-infra[ai]

With API Integration

For REST/WebSocket APIs using svc-infra:

pip install robo-infra[api]

With Hardware Support

For real hardware buses (I2C, SPI, Serial, CAN):

pip install robo-infra[hardware]

Full Installation

Install everything:

pip install robo-infra[full]

Platform-Specific Bundles

Platform	Command
Raspberry Pi	pip install robo-infra[raspberry-pi]
NVIDIA Jetson	pip install robo-infra[jetson]
Generic Linux	pip install robo-infra[hardware]

Granular Hardware Extras

Feature	Extra	Command
I2C sensors	[i2c]	pip install robo-infra[i2c]
SPI devices	[spi]	pip install robo-infra[spi]
Serial/UART	[serial]	pip install robo-infra[serial]
CAN bus	[can]	pip install robo-infra[can]
GPIO	[gpio]	pip install robo-infra[gpio]

Vision & Camera Extras

Feature	Extra	Command
OpenCV	[vision]	pip install robo-infra[vision]
Intel RealSense	[realsense]	pip install robo-infra[realsense]
Luxonis OAK	[oak]	pip install robo-infra[oak]
Raspberry Pi Camera	[picamera]	pip install robo-infra[picamera]
All cameras	[cameras]	pip install robo-infra[cameras]

Quick Start

from robo_infra.actuators import Servo
from robo_infra.core.types import Limits

# Create a servo (works in simulation by default)
servo = Servo(
    name="gripper",
    channel=0,
    limits=Limits(min_value=0, max_value=180),
)

# Move to position
await servo.move_to(90)

# Get current position
print(f"Position: {servo.position}°")

With Real Hardware

from robo_infra.drivers import PCA9685Driver
from robo_infra.actuators import Servo

# Initialize hardware driver
driver = PCA9685Driver(i2c_address=0x40)

# Attach servo to driver
servo = Servo(
    name="gripper",
    driver=driver,
    channel=0,
)

await servo.move_to(90)

AI Integration

from ai_infra import Agent
from robo_infra.actuators import Servo
from robo_infra.integrations import RobotTools

# Create robot components
servo = Servo(name="arm", channel=0)

# Export as AI tools
tools = RobotTools([servo])

# Use with AI agent
agent = Agent(tools=tools.as_tools())
await agent.run("Move the arm to 45 degrees")

API Integration

from svc_infra import create_app
from robo_infra.actuators import Servo
from robo_infra.integrations import RobotRouter

# Create robot components
servo = Servo(name="arm", channel=0)

# Export as API router
router = RobotRouter([servo])

# Add to app
app = create_app()
app.include_router(router.as_router(), prefix="/robot")

Core Concepts

Actuators

Physical components that create movement:

• Servo - Position-controlled rotational actuators
• Motor - Speed-controlled rotational actuators
• Stepper - Precise step-based motors
• LinearActuator - Linear motion actuators

Sensors

Components that measure the environment:

• IMU - Inertial measurement (accelerometer, gyroscope)
• Distance - Distance measurement (ultrasonic, ToF, IR)
• Temperature - Temperature sensors
• Current - Current sensing for motor feedback

Drivers

Hardware interfaces:

• PCA9685 - 16-channel PWM driver
• ADS1115 - 4-channel ADC
• MCP23017 - 16-bit I/O expander

Controllers

High-level control systems:

• ArmController - Robotic arm coordination
• DriveController - Differential/holonomic drive
• FlightController - Quadcopter/drone control

Documentation

• Getting Started
• Hardware Setup
• API Reference
• Examples

Part of the nfrax Ecosystem

robo-infra is designed to work seamlessly with:

• svc-infra - Backend infrastructure (API, auth, database)
• ai-infra - AI/LLM infrastructure (agents, tools, embeddings)

License

MIT License - see LICENSE for details.