Future Work
There are numerous applications of this system. I have listed a few which can be developed in the Tekkotsu environment by building on this program.
| New Final Position | Code could be added to allow the user to define where the arm should move the target once it has been grabbed. This functionality could lead to more applications some of which are desribed in this table. |
| Tower of Hanoi | One of the original goals for this project which was later deemed out of scope. The program could be adapted so that the camera would recognize three or four blocks in a row. It would use the arm to solve the Tower of Hanoi puzzle using the blocks. This would need to be combined with the "New Final Position" goal in order to be functional. |
| Other half of the Table | This new feature came from an assumption made early in the project to simplify it. As I started developing this program I only considered one half of the table. I had intended to adapt the code to handle both sides equally but ran out of time. The code would need to mirror the arm movements to handle the other half of the table. As the core functionality existed for at least half of the table it was deemed a new feature to be added at a later date. |
| Pathfinding with cognition | The ultimate form of this base functionality would work in a very "cluttered" environment. That is, there could be numerous objects which the arm might hit if it does not acount for them when moving. The current system moves the arm to an assumed-safe starting position before attempting to grab the target. The ultimate version of this program could identify objects in the way of that motion and acount for them. However there is another issue. The arm currently lines up with the target, moves towards it obscurely and finally grabs the target. The program should be able to move from wherever the arm may be when the program starts smoothly to the target directly while avoiding obstacles as described above. |