SoftBank Robotics documentation What's new in NAOqi 2.8?

Python - How to write a qimessaging client

Introduction

This guide will teach you how to write a qimessaging client in Python, to interact with the various services offered by the robot.

Prerequisites

  • An installed python NAOqi SDK for your operating system.

Creating an application

The first step to interact with qimessaging services is to create an Application, it will handle the connection to the robot.

import qi

app = qi.Application()
app.start()
session = app.session

Calling a method on a service

Let us make the robot speak by using the say method of the ALTextToSpeech service.

tts = session.service("ALTextToSpeech")
tts.say("Hello Word")

Starting your application

You can then run your application. It will connect to localhost by default, but you can override that.

python MyApp.py --qi-url tcp://nao.local:9559

For more information on the arguments your application can take, see: qi application arguments.

Making asynchronous calls using qi::async

Most of the API are designed to block until the operation started by the call completes. For example with this code inside our main:

tts = session.service("ALTextToSpeech")
motion = session.service("ALMotion")
tts.say("This is a very very very very long sentence.")
motion.moveTo(1, 0, 0); // go forward one meter

The robot will only start moving when he finishes speaking.

To perform both actions simultaneously, the API provides a special _async argument that performs a call in an asynchronous manner, and notifies you when the call finishes using a qi.Future:

tts = session.service("ALTextToSpeech")
motion = session.service("ALMotion")
sayOp = tts.say("This is a very very very very long sentence.", _async=True)
moveOp = motion.moveTo(1, 0, 0, _async=True)
# Wait for both operations to terminate.
sayOp.wait()
moveOp.wait()

Look at the qi.Future for more complete documentation, but here is what you most definitely need to know:

  • If the method throws an exception, it is stored in the qi.Future, and can be accessed using qi.Future.error().
  • Use qi.Future.wait() to wait for the future to complete. It can accept a timeout duration as argument, and will return the state of the future.
  • Use qi.Future.value() and qi.Future.error() to get the stored value or error.
  • You can register a callback to be notified when the future finishes with qi.Future.add_callback().