With the use of NLP multi-intent classifier, the robot is able to understand the user's command and execute the corresponding task. Steps of the task includes: listening to the user's command generated by a program by Robocup. Acknowledge the command by saying it back to the user. Executre the task. Report the result to the user.
Using preprogramed actions per intent as modules, then ssc_generator.py takes in the multi-intent split them, and generate a smach state machine. the generated ssm is saved as temp_generated_smach_state.py. The state machine is then is executed by immediately by the generator as a seperate thread.
To start run
# fixed parameter, great for testing
python task_gpsr_manual.py
# using NLP to extract multi-intent
python task_gpsr_auto_gen.pyAuthor: Tinapat Limsila