getting-started-with-python.md

Connecting to the simulator with Python

You can use the python client to connect the simulator. The python client is able to retrieve sensordata and send vehicle controll.

Dependencies

The Python client depends on msgpack, numpy and opencv-contrib. Install the dependencies like this:

pip install -r requirements.txt

Getting started

Let's drive the car forward!

# This code adds the fsds package to the pyhthon path.
# It assumes the fsds repo is cloned in the home directory.
# Replace fsds_lib_path with a path to wherever the python directory is located.
import sys, os
fsds_lib_path = os.path.join(os.path.expanduser("~"), "Formula-Student-Driverless-Simulator", "python")
sys.path.insert(0, fsds_lib_path)

import time
import fsds

# connect to the AirSim simulator 
client = fsds.FSDSClient()

# Check network connection
client.confirmConnection()

# After enabling api controll only the api can controll the car. 
# Direct keyboard and joystick into the simulator are disabled.
# If you want to still be able to drive with the keyboard while also 
# controll the car using the api, call client.enableApiControl(False)
client.enableApiControl(True)

# Instruct the car to go full-speed forward
car_controls = fsds.CarControls()
car_controls.throttle = 1
client.setCarControls(car_controls)

time.sleep(5)

# Places the vehicle back at it's original position
client.reset()

A full example of an autonomous system that can finish a lap can be found here

Find more examples here.

Sensors

Documentation on requesting and processing sensordata can be found in the respective sensor documentation pages:

• Lidar
• Camera
• GPS
• IMU
• Ground Speed Sensor

Getting ground truth information

Using the following function you get get the real, latest position of the car:

state = client.getCarState()

# velocity in m/s in the car's reference frame
print(state.speed)

# nanosecond timestamp of the latest physics update
print(state.timestamp)

# position (meter) in global reference frame.
print(state.kinematics_estimated.position)

# orientation (Quaternionr) in global reference frame.
print(state.kinematics_estimated.orientation)

# m/s
print(state.kinematics_estimated.linear_velocity)

# rad/s
print(state.kinematics_estimated.angular_velocity)

# m/s^2
print(state.kinematics_estimated.linear_acceleration)

# rad/s^2
print(state.kinematics_estimated.angular_acceleration)