PHRI1920_dev

Introduction

1. We have two NICO robots: A and B
2. Each robot can have LEFT or RIGHT position. We use the robot's perspective, not a human. LEFT position means that a robot on the right side of a commander
3. See markings on the floor for placing robots and a table correctly
4. Each robot runs on its own computer. You need sudo access
5. LEFT robot runs on LEFT computer (operator's perspective)

Pre-check

1. Make sure that both robots are connected to the power source. Motors in a relaxed state produce noise, you can here it. One NICO needs two power sockets and the other only one.
2. Connect all USBs and audio cables to PCs.
   1. Check if faces are lighted up
   2. Check if speakers on. Check the sound settings on a PC.

Initial Set-up

A general set-up process for two computers

1. git clone https://git.informatik.uni-hamburg.de/wtm-robots-and-equipment/NICO-software.git
2. cd NICO-software/api
3. source NICO-setup.bash
4. cd NICO-software/api/src
5. git clone https://git.informatik.uni-hamburg.de/wtm-teaching-projects/phri1920_dev.git
6. cd NICO-software/api
7. catkin_make
8. source devel/setup.bash
9. sudo chmod 777 /dev/ttyACM*
10. sudo chmod 777 /dev/ttyUSB*

A set-up for LEFT computer

1. On this computer run 4 services and one subscriber
   1. pose service
   2. face expression subscriber
   3. speech synthesis service
   4. speech recognition service
   5. vision service for counting cubes
2. For each service/subscriber a new terminal is needed. The set-up process for each terminal:
   1. export ROS_MASTER_URI=http://wtmpc23:11311/
   2. source ~/.NICO/bin/activate
   3. cd NICO-software/api
   4. source devel/setup.bash
3. Start services:
   1. rosrun nicopose poseService.py --label A --position LEFT
   2. rosrun nicopose fexSub.py --label A --position LEFT
   3. rosrun speech SpeechSynthesisStub.py A
   4. rosrun speech SpeechRecognitionStub.py
   5. rosrun vision count_resources.py

A set-up for RIGHT computer

1. On this computer run 4 services and one subscriber
   1. pose service
   2. face expression subscriber
   3. speech synthesis service
   4. state machine
2. For each service/subscriber a new terminal is needed. The set-up process for each terminal:
   1. export ROS_MASTER_URI=http://wtmpc23:11311/
   2. source ~/.NICO/bin/activate
   3. cd NICO-software/api
   4. source devel/setup.bash
3. Start services:
   1. rosrun nicopose poseService.py --label B --position RIGHT
   2. rosrun nicopose fexSub.py --label B --position RIGHT
   3. rosrun speech SpeechSynthesisStub.py B
   4. rosrun state_machine StateMachine.py --scene 0 --entry 0

Ros launch scripts

You can use launch scripts to avoid a lot of typing, which is described in the previous section.

1. Launch nicopose roslaunch nicopose pose.launch llabel:=A lnvc:=M
2. Launch logging: rostopic echo /rosout | grep msg

Experiment set-up.

Note, you need to change lnvc parameter according to your setup. Start terminal. In each terminal session execute the required commands:

1. Computer LEFT:
   1. Launch nicopose roslaunch nicopose pose.launch llabel:=A lnvc:=M
   2. Launch logging: rostopic echo /rosout | grep msg
   3. Execute rosrun vision count_resources.py
2. Computer RIGHT:
   1. Launch nicopose roslaunch nicopose pose.launch llabel:=B lnvc:=S
   2. Execute rosrun speech SpeechRecognitionStub.py
   3. Execute rosservice call speech_recognition calibrate
3. Computer iCub. Note, you must run cd /data/home/hri/phri1920 and source setup.bash in each session
   1. Execute roscore
   2. Execute roslaunch state_machine lab_pc.launch before participant enters
   3. Execute rosrun state_machine StateMachine.py when robots are supposed to start talking.
      • optional parameters are e.g. : --scene=0 --entry=0
   4. You can set the lights manually, if needed: rosservice call /light_control "setting: 'cyan'"

NB: In case of face expression is not recognized for B (right PC) open a terminal

1. check which ttyACM "ls /dev/ttyACM" and tab
2. go to fexSub.py change specifically line 37, change to respective ttyACM

Links

PHRI1920 Wiki