Locations, objects and questions referred in the description of the tasks below always refer to one among the set of predefined items.
The home position (i.e., starting pose of the robot for each test) and other three locations will be defined on site with name and position in the environment (e.g., kitchen, living room, table).
A set of per-defined objects will be communicated soon here.
1. When RoboCup@Home started?
2. In which city are we now?
3. What is the name of your team?
4. Who lives in a pineapple under the sea?
5. What is the highest mountain in Europe?
6. Who invented the telephone?
Please prepare appropriate answers. If you don't know the answer, find it out.
Setup - Robot starts in the home position, a person is waiting in one locations (known to the robot)
Phase 1 - (navigation) - Robot reaches a predefined known location
Phase 2 - (speech synthesis) - Robot says that it is ready to follow the person
Phase 3 - (person following) - Person moves slowly in a direction and robot has to follow him/her to an unknown destination
Phase 4 - (speech understanding) - Person places an object on the robot and issues a spoken command describing the location where the object must be delivered
Phase 5 - (navigation) - Robot reaches the given location
Phase 6 - (speech synthesis) - Robot announces that it is arrived to the destination
Setup - Robot facing one direction, one person holding one object is placed behind it in unknown position (about 1.5 meters away from the robot) - no other obstacles will be close to the person.
Phase 1 - (navigation) - Robot turns around
Phase 2 - (people/face detection) - Robot faces the person, and gets closer to him/her (e.g., 0.75 m)
Phase 3 - (speech synthesis) - Robot greets the person
Phase 4 - (speech understanding) - Person asks a question to the robot and the robot has to answer properly
Phase 5 - (object recognition) - Person shows an object to the robot and the robot has to recognize the object and to explicitly say its name.
Setup - Robot starts in home position, a person is at a known location and two objects (unknown to the robot) are placed in another location (unknown to the robot).
Phase 1 - (navigation) - Robot reaches the person
Phase 2 - (people/face detection) - Robot faces the person, and gets closer to him/her (e.g., 0.75 m)
Phase 3 - (speech synthesis) - Robot asks if s/he wants something
Phase 4 - (speech understanding) - Person answers that s/he wants an object and the robot repeats the object name
Phase 5 - (speech synthesis) - Robot asks where the object is
Phase 6 - (speech understanding) - Person answers the location of the object and the robot repeats the location
Phase 7 - (navigation) - Robot goes to location of the object where two objects are present (at least 1 m away each other)
Phase 8 - (object recognition) - Robot finds the correct objects facing it and clearly declaring its name; if object is correctly recognized, the referee will place it on the robot
Phase 9 - (navigation) - Robot goes back to person and reports which object was found
Bypass rules can be applied to bypass any functionality (i.e., any phase of a test).
Score for that phase will be zero, but it is possible to gain score for next phases.
Navigation and person phases can be bypassed with the use of a joystick or requiring the person to move towards the robot and continue with the test in the current position of the robot
Speech understanding can be bypassed by choosing an appropriate element (e.g. object or location) instead of the one spoken by the person. The chosen element must be declared by the robot with a spoken sentence or by the team leader.
Speech synthesis can be bypassed by writing text on a screen (possibly on board the robot)
that must be clearly shown to referees by the team leader.
Object recognition can be bypassed by declaring that the object cannot be recognized and skipping the phase.
People/face detection can be bypassed by asking the person to get in front of the robot.
The total score is computed as the sum of points achieved for correctly completing any phase in any task. However, the score for each phase depends on previous achievements of the phases with the same functionality.
More specifically,
1st achievement of a functionality -> 100 points
2nd achievement of a functionality -> 100 -20% = 80 points
3nd achievement of a functionality -> 80 -20% = 64 points
and so on ...
Repeating the same functionality many times gives less score than succeeding in many functionalities.
Example:
Team A performs 3 navigation phases --> score = 100 + 80 + 64 = 244
Team B performs 1 navigation, 1 speech synthesis, 1 speech recognition phases --> score = 100 + 100 + 100 = 300