Demonstration Day at Cornell University
In our last days of working together as a group in person, we demonstrated our results with other research teams in the department. The videos below show our demonstration, annotated slightly for clarity. The situation is meant to simulate the youBot finding a red object then closing in on it to determine if it can or cannot pick up the object on its own. For the sake of the demonstration, the youBot was hardcoded to believe that it could not reach the object on its own due to an obstacle, and the obstacle (such as a small table or ledge) was not including in the demonstration for ease of filming and seeing exactly what was going on with the minion bot.
|
(If your computer can handle it, the above video is best viewed in 1080p for full effect) |
As you can see at the end of the video, the youBot believes that the minion bot is on top of the red object prematurely and commands it to go into its grab function. This is most likely the product of inaccuracies in the distance sensor of the camera on the youBot, and could most likely be mitigated in future iterations if they were to occur.
To demonstrate that the minion bot can in fact grab hold of the intended object, we then placed the minion bot on the object and instructed it to go into its grab function. The below video shows the result.
The remaining task would be to get the minion bot back to the youBot to be picked up again. It is possible that the remaining two legs of the minion bot would be enough to shuffle the bot back out from the confined area, at least far enough to be in the reach of the youBot. If this were not to be the case, then it is possible that the youBot could draw in the minion bot directly using something like a winch which would pull on the tether. Then once the minion bot is close enough to the youBot it can be picked up again. These are efforts we would pursue in future iterations that do seem generally feasible.