The bottom line: what most respondents seemed to look for, after aggregating all the answers, were ease, speed, convenience, consistency, and water taste.

In general, the water filling stations are pretty good as they are, and certainly are better than the old method of refilling water bottles from normal drinking fountains. Most users were satisfied with their experience. However, there could always be room for improvement. I have some suggestions that can make the overall process and experience more user-friendly.

Combining both the manual and automatic types seems to be the best solution that could get rid of many of the problems that users faced. One solution is to have the water filling station be automatic to start, activated by the sensor, but the user can press a button to stop it. Another solution is setting it so that pressing the button the first time turns the station on, and pressing it again turns the station off, like a simple switch. Both solutions keep the station hands-free while eliminating the problem of overfilling the water bottle and inconveniencing the user upon removal. Reducing the need for a motion sensor, like in the second proposed solution, could also cut down cost.

In addition, it would be ideal to have a consistent type of water filling station all throughout campus so users would immediately know what to do instead of having to waste time figuring out what type of station it is. I'd keep the interface for this station as simple as possible. I tested the water stations and for most of the stations it is possible to drink at the same time that a water bottle is being filled, but as seen from user responses a lot of users feel awkward about that. Thus, at least 50% of the current interface is always out of use, because users typically don't drink while someone else is filling a water bottle and vice versa. Because of this, I'd separate the water bottle filling station and the water fountain part of the interface and place them side by side, as is the case with typical drinking fountains.

Finally, the last major thing that users commented on was the ambiguity of the filter display. As it is, the display is only visible when users are standing in front of the station. Instead of putting the display in the top band, I would have a light on the side of the station to show whether or not a filter is in need of changing. This way, users will already know when approaching the station if the filter is out-of-date, and the service workers who replace the filters can more easily see which fountains are in need of changing filters. The current display has three states, green, yellow, and red, and it is unclear for users to know what exactly yellow means (and whether it is good to drink). I'd eliminate that intermediate state completely.