1. The NAO
Soccer Robots Implementing soccer behaviors in the Webots simulator for NAO robots. Kristjan Arumae, Sarah Myhre, Chris Cassian Olschewski
Sponsor: Dr. Gita Sukthankar
Adviser: Astrid Jackson
Senior Design: Dr. Mark Heinrich

2. Project Objectives
Create autonomous agents with soccer behaviors in a simulation environment
Create resources to guide future students working with the NAO simulator or robots

3. Current State Soccer simulation code is rare
Most code develop for SimSpark
Complexity through years of iteration
Documentation limited and incomplete
Difficult to adopt for own purposes, not designed for flexibility

4. Solutions by (Senior) Design
Make code public and easily available
Develop in Webots for NAO
Structure low-level design in a modular, easy to approach way
Have thorough and complete documentation
Design code for multiple functions

5. Major Components Robot Locomotion & Decision Making Online Resources
Robot Vision

6. Team Development Windows, Mac, Linux Linux Java Python
Choreographe                        Naoqi API
Webots Trackers                    OpenCV

7. Technical Design

8. Major Components Robot Locomotion & Decision Making Online Resources
Robot Vision

9. Robot Vision: OpenCV
RGB
HSV
BGR
Grayscale

10. Robot Vision: In Code

11. Robot Vision: In Code

12. Robot Vision: Bounding Boxes

13. Major Components Robot Locomotion & Decision Making Online Resources
Robot Vision

14. NAOqi: Anatomy of a Kick
Legs fixed to constrains
Balance activated for support leg
Duration for balance leg mode set
3 lists of actuator settings passed
3 time steps passed and motion executed

15. NAOqi: Kick Locomotion

16. NAOqi: Decision Making
NAO Behavior Tree
Goalie

17. NAOqi: Decision Making
Behavior
→ init stance
→ find ball
→ block

18. NAOqi: Decision Making
NAO Behavior Tree
Striker

19. NAOqi: Decision Making
Striker Behavior Walk Towards

20. NAOqi: Decision Making

21. Major Components Robot Locomotion & Decision Making Online Resources
Robot Vision

22. Resource: Online Wiki

23. Future Applications Memory of World State, Confidences
Artificial Neural Networks
Locomotion Training with Genetic Algorithms

24. Implements proposed solutions?
Evaluation
Implements proposed solutions?

25. Implements proposed solutions!
Evaluation
Implements proposed solutions! ü
Widely available on public git repository:

26. Implements proposed solutions!
Evaluation
Implements proposed solutions! ü
Widely available on public git repository: ü
Implementation specifically tailored to Webots

27. Implements proposed solutions!
Evaluation
Implements proposed solutions! ü
Widely available on public git repository: ü
Implementation specifically tailored to Webots ü
Code with different levels of sophistication

28. Implements proposed solutions!
Evaluation
Implements proposed solutions! ü
Widely available on public git repository: ü
Implementation specifically tailored to Webots ü
Code with different levels of sophistication ü
All documentation online: NAO UCF website

29. Implements proposed solutions!
Evaluation
Implements proposed solutions! ü
Widely available on public git repository: ü
Implementation specifically tailored to Webots ü
Code with different levels of sophistication ü ü
All documentation online: NAO UCF website ü
Team templates and plug’n’play BTs/FSMs

30. CS Senior Design Year 1 Complete
Team Soccer Robots created autonomous agents with soccer behaviors in a simulation environment
Team Soccer Robots created resources to guide future students working with the NAO simulator or robots