Progress so far

Created 2D calibration method using haptic feedback of the touchscreen. (spring 2025)
Gathered accuracy results of reach movement with this calibration (correction) used. (summer 2025)
Compared new results (correction used) with old ones (no correction used). (summer 2025)
Objective 1 finished and published in ICANN 2025. (summer 2025)

Managed to enable grasping in all available positions using palm orientation interpolation. The palm orientation (yaw) must change based on the position; with a fixed orientation, the robot cannot reach everywhere. (autumn 2025)
Managed to detect large ball using YOLO algorithm and consequently get the 2D position of the ball just from this visual input and kinematic chain. (autumn 2025)
Created script that finds the ball using stereo camera in robot's eyes, gets the position of the ball and attempts to grab it. (autumn 2025)
This ball is however not suitable for NICO grasping (too large), it needs smaller objects. YOLO algorithm was not able to detect smaller objects.

Wrote the first 15 pages of the thesis containg introduction and theory. (winter 2025)

Future work

Find appropriate object detection model for finding objects that NICO can grasp.
Create and realize testing of this method that finds and grasps objects.

Find a solution to finding 3D position of an object using just a visual input.
Create a script that finds an object in any 3D position and attempts to grasp it.
Create and realize testing of this method that finds and grasps objects in 3D.