1 hour lecture + 2 hour exercises at computer

1. Action. Actuators. The robot gear control: via position, via speed. The robot arm control. Direct and inverse kinematics.
Exercise: iCubSim touching ball on table

2. Perception. Sensors. The basic processing of sensor data: distance, camera image, depth map, tactile map.
Exercise: iCubSim seeing ball on table via color detection

3. Control. Decomposition of the control system by function and by activity. Significance of real time.
Exercise: simulated ALLEN exploring room

4. The regular objects recognition. Hough transform.
Exercise: Boebot following ping-pong ball

5. The irregular objects recognition: DOT/HOG, SIFT/SURF/ORB, phase correlation. Representation of objects in robot memory. Sensory-motor approach to perception.
Exercise: Boebot following selected object via phase correlation and via SURF.

6. Behavioral robotics. Emergence of control in modular control architecture.
Exercise: Simulated ALLEN combining selectable behaviors.

7. Cognitive approach to robot control. GOFAI, planning. STRIPS. Sussman's anomaly. Frame problem
Exercise: Simulated SHRDLU manipulating cubes

8. Post-cognitive approach to robot control. Dreyfus' criticism of GOFAI. Brooks' subsumption architecture. Situated robots. Embodiment. Interaction principle (robot ACE).
Exercise: Simulated ALLEN, Brooks' subsumption

9. Minsky' society model of mind. Inspiration from Piaget's developmental psychology. Dennet's mind types. Robot COG.
Exercise: iCubSim interacting with human

10. Intelligence as a social phenomenon in group of robots.
Exercise: SWARM.

11. Robots recognizing and emulating emotions. Robot KISMET.
Exercise: Courting between robots according Freud's pleasure pump theory.

12. Cloud technology for robots. Robot Pepper. IBM Watson. MicroSoft Azure.
Exercise: Simulated robots coordinated via web service.

13. Control emerging from interaction of robot with its environment. Delayed reinforcement learning, Genetic programing, Neural network training.
Exercise: Control derived from data via deep learning