1. 10th International DAAAM Baltic Conference "INDUSTRIAL ENGINEERING”
12­13th May 2015, Tallinn, Estonia
Matti Mielonen
Joni Väisänen
Olli Hasu
Toni Hevonoja

2. Background Evolution of human anatomy Previous research
Fine and complicated mechanisms
Artificial imitation is interesting and challenging
Previous research

3. Scope of the research
Can human anatomy be effectively imitated by pneumatic muscles?
Application
Sports training robot (“pneumatic athlete”)
Primary goals
Pneumatic human anatomy imitation (in sport context)
Utilization of pneumatic muscles
High performance and accuracy

4. Football kicking device

5. Mechanical design

6. Pneumatic systems

7. Electronics and software
RASPBERRY PI 2
ARDUINIO
NANO

8. Results Performance evaluation Kicking action and velocity Accuracy
High speed camera
Accuracy
Visual evaluation of goal scatter (parameters unchanged)
Accuracy of control system
Mounted laser pointer to help visual evaluation

9. Results: kick action

10. Results: kick action

11. Results: velocity

12. Results: kick action Performance evaluation Kicking speed
Achieved: km/h (static)
Target speed: 100 km/h
Not a pure kick (2 impacts due to muscle suspension)

13. Results: accuracy

14. Results: accuracy Performance evaluation Accuracy Mechanical accuracy
Impacts inside a circle of 1m radius
Accuracy of control systems
Alignment: unstable and inaccurate
Ankle: stable and accurate

15. Further research (“Big Picture”)
Focus
Direction of development
Current state
Conceptual prototype
Sport training application
Increase of performance
Supportive features
(machine vision)
Human anatomy research
Finer and more realistic imitation
and / or

16. Conclusion
High-performance pneumatic imitation of human anatomy by fluidic muscles
Possible, yet further refinement is needed
Mechanical improvements (transmission ratio, precision, more realistic imitation…)
Accuracy of control systems
Questions raised:
How to improve performance and accuracy?
How to achieve more realistic imitation?
Potential for commercial use?
More features (i.e. machine vision)

17. Thank you!