This image shows the knob for a radiator in a building on MIT campus. This knob implementation has two major problems for dimensions of usability. The first is with learnability; the knob has no clear indicator for whether larger or smaller numbers correspond with more or less heat emitted from the radiator. As a result, users will not be able to look at this knob and immediately discern how to use it to change the radiator's temperature. This is an example of poor learnability. The second major issue is along the dimension of safety. Temperature is very sensitive and in some cases may result in damage to personal health or objects which are either valuable or volatile. Since this knob leaves open the possibility that users turn it to a much hotter or colder setting than intended, there is a safety flaw.
This knob could be improved by including a simple temperature indicator alongside the numbers that circumscribe the knob. For instance, there could be a line that begins with red color (at the end which corresponds to the warmest temperature setting), ends with blue color (at the end with the coldest temperature setting) and follows a gradient in between.
This image shows a very short table intended for use in a room. This table fails along the dimension of efficiency first and foremost. It is roughly 8 inches off the ground (a leather couch is visible in the background, and its seat is considerably higher than the table). For this reason, the table is incredibly inefficient for any uses that don't involve setting an object on a surface above ground level. Any tasks that could be done at a table of typical height (i.e. a height that is usable for someone who is either standing or in a chair) are not feasible with this table, which is far too short. This lack of usability from a seated or standing position is compounded by the fact that the table is entirely circular; the complete lack of concavity on the top surface's perimeter means that a user sitting in front of the table has a very limited amount of surface area to work with before they have to lean further into the table to use its space.
If this table were being redesigned to increase usability and efficiency while still maintaining some of its primary stylistic choices (curvature, slender and sleek frame), it could be improved by slightly raising it to a level at which it is accessible to someone sitting on a low couch without demanding that the user lean forward so as to be leaning off the couch. Additionally, one portion of the circular surface could be replaced by a smooth, concave inlet that allows one person to sit in front of the table and be surrounded on 3 sides by the table's surface. These changes would improve the table's efficiency.
This image shows a uniquely-designed bookshelf that lets books rest at off-angles along a vertical axis. This bookshelf has issues with all three dimensions of usability. First, it has a problem with learnability, given that users who are placing books on the bookshelf when it is empty cannot readily tell which direction the books should be placed in. Second, it has a problem with safety. This is because the lopsided nature of the shelves makes it so that a user can place all their books on only the left or right side of the shelf, and when this occurs, the shelf topples over. Additionally, stacking books against one another with uneven distribution of forces can be damaging to books further down the stack, which is harmful to the products that this shelf is designed for. Third, it has a problem with efficiency. The safety and learnability issues demand that users struggle with deciding where to place the books, and the lopsided nature of the shelf makes it so that users likely will need to use some trial and error to determine where to place books.
Many of the problems with this bookshelf demand changes that reject some of the style choices used here. Most notably, the diagonal orientation is simply unsafe for storing books, and should be replaced with typical right-angle shelving in order to increase safety. This will naturally give users more learnability, since the shelf will be in a familiar orientation, reducing any confusion that users may have when placing books on the shelf for the first time. Additionally, this could be improved by adding some way to more evenly distribute weight across the right and left sides of the shelf. This might come from extending the shelves out so that they have equal lengths on either side of the shelf's sagittal plane, or it may come from a clever counterbalancing mechanism that redistributes weight in the shelf in response to increases/decreases of weight on one side.