| links | |
|
[home] [high level design] [task distribution] [hardware design] [software design] [design results] [conclusions] [media] [links] |
The Automated Beverage DispenserTeam Members:Michael Everhart David Castellani Design Results: 1) Functionality: Upon final completion of our design, following a grueling debugging process, the design worked very close to initially intended. A cup was placed into the device – the device would move itself into a pouring position, and then water would begin to pour. Due to the reaction capacity of the servo controlling the cup angle, we were not able to steadily lower the cup but instead had to rather quickly bring it to different levels along the pouring arc. 2) Device Constraints: One device constraint we had to deal with was the behavior of the sonic ranger. The sonic ranger was both consistent and accurate when used in an open environment, however, it had much more difficulty when it was pointed into the cup. We suspect that the sound waves echoed and bounced around at random angles inside the cup, leading to inconsistent results. In the final design, we ended up moving the sonic ranger much closer to the mouth of the cup than we had originally intended. This cleared up most of the issues with the sonic ranger, but it also was more likely to be bumped while inserting or removing the cup. 3) Control Implementation Issues: Control implementation had to be changed in a minor way on a number of devices due to the difference between specification usage and implemented usage. The most prolific example of this was the problems faced with the sonic ranger. The ranger seemed to act quite sporadically when used in conjunction with the rest of the modules on the device as compared to the way it acted when it was used in a more controlled environment. We attempted to remedy the problem in many ways including reducing the number of steps that the servo took to bring the cup to an upright position and reimplementing some hardware features. The eventual problem rested with noise on the line of the device, however the solution was found late and the slight limitations on the functionality were left in place. 4) Unresolved Problems: By the time the design was finished and presented, the only unresolved problems rested with the sonic ranger. Given a bit more time to manipulate the sensor we would have most likely been able to remedy the problems. The major unresolved problem is that on an occasion the cup will be moving at the same time that the ranger attempts to send a sonic ping. In some cases, this results in the ping being reflected off of the wrong part of the cup, leading the software to believe that the liquid level is much high than it actually is. This causes the software to move the cup to the next most upright position prematurely. All material copyright 2005 Michael Everhart and David Castellani. |