In order to test the framework, some experiments will be done.
All experiments are done in the unity side of the framework as I dont have access to the actual robot at the moment.
In order to explore the capabilities of this implemented design, experiments were conducted to test the following criteria:
The experiments include a setup that includes 3 components: (1) The Robot Model, which will do all the tasks; (2) the Target Object, which will be what the model will pick up; and (3) the Target Placement, which is the location in which the robot will place the target object. As seen above, the Robot is placed in the center of a 3D table asset, which has a rigid body property (same as all other components). This is so that the measurement of the range of the model can easily be made. In the designed experiment, the Target Object’s starting coordinates will be considered the independent variable. This means that the starting point will vary per experiment, which will help give insight on the aforementioned criteria. For the Target Placement, it will stay in the same position all throughout the experiment.
Since the experiments are in Unity3D, it follows the coordinate system and scale of that game engine. The software’s coordinates are seen in the figure.
Viewing the simulation from the XZ plane, the workable range of the robot model is in the range of -0.3 to 0.3, both in the X and Z axes, as illustrated by the gridlines. Using this range, it will define the possible coordinates of where the target objects will be placed.
Knowing the possible coordinates, the flow of the experiment can be easily defined. As seen in the figure below, each iteration of the experiment starts with a specific coordinate of the Target Object. After the object is set, a pick and place script will be run, in which case the performance of the robot will be observed and assessed. After the target object is successfully or unsuccessfully picked up, a new iteration of the experiment will start but with a new coordinate for the target object. The order in which the coordinates will be set is from [-0.3,-0.3] to [0.3,0.3] moving at 0.1 increments until all possible areas in the grid have been tested. At each coordinate point, it will be marked whether a trajectory was possible, and if the pickup was successful or not. Along with this, factors regarding the design of the arm are also observed. After all the experiments are done, the data will be recorded and plotted.