Our final consists of 2 main parts and the shelter as a supporting element. The floor component is for the most basic bus stops, the ones without a shelter or seat. Used alone it serves as a visual guide for patrons to queue. With the seat, they create a gentle barrier to prevent the majority of queue-cutting while also providing comfort.

While only a limited number will queue at the start, patrons chose to do so of their own accord as it places them closer to the boarding area. Our designs do not force patrons to queue while waiting, but queuing comes into effect once the bus is in visible sight – seated patrons closest to the door are now able to board first, other patrons follow the floor guides to queue with only a small number who might disregard the guides cutting in. We believe this design achieves our goal to guide international patrons by creating a local queuing culture that is intuitive for all patrons, while doing it in a gentle way.


– Clear visual guide.

– Colour chosen for its visibility and link to bus stop.

– Linear for easy installation and adaptable shape.

– Pointed circle serves as initial focal point to guide the eye.

– Granulated rubber.

– Low cost.

– Easy installation.

– Nice texture and grip.

– Good colour.

– Made from recycled materials.

– Angled for better view for bus for all patrons.


– Serves as gentle barrier between the seat and pole. (Reduces the number of patrons who cut in)

– Only 2 different pieces to reduce need for many moulds.

– Seats an extra person in the same amount of space.

– Angled for better line of sight of bus.

– Concrete base for strength, aesthetic and low-cost.

– Wooden top for comfort, familiarity and contrast with concrete.

– Modular to allow for different bus stop capacity and shapes.

Shelter (Supporting Element)

– Show how a shelter can work with the other 2 levels.

– Gradient of coverage, those who arrived first gets the best comfort (seat and shelter).

– Opportunity for funding through advertisement.


The result is a modular installation with 3 levels of service. It is easy to install, and can be easily adapted to fit many situations and needs:



Scenario Testing

Once we had the core of our final established, we decided to do a scenario test. Marking out an area identical to our example bus stop, we decided to roleplay as patrons.

Scenario Testing

Drawing from the knowledge we gained in field testing we could easily see what the worst case and best case scenario would be. The best case would be that patrons start lining up orderly even before the bus arrives, once the bus comes, seated patrons (usually those who arrive at the stop first) would be closest to the door and board first, other would file orderly along the floor marking. Worst case scenario would simply be that users disregard the installation, or intentionally cut the queue (eg. climbing over the seat or walking around the front of the bus pole).



I think the greatest advantage of this method is that it can give a similar result to field testing without the need to observe strangers. The disadvantage is that it can limit the outcomes a little. For example, in the field test, we saw a patron begin to cut in front of the bus pole, but upon seeing the queue, walked the full way around the bench to line up. Or for situation where more than 1 bus arrives at the same time, or bus stopping in the wrong area. Overall I think field testing is better, but when that is not always an option, it scenario testing is a great accompaniment or fair substitute.

Mapping patron’s BEFORE our installation (When waiting for bus vs when bus arrives):

Mapping patron’s AFTER our installation (When waiting for bus vs when bus arrives):


Field Testing


We chose to do our field at a Mt Eden bus stop as Ben had previously done observations on this stop so that we would have a baseline for compassion, additionally it is a stop that has heavy rush hour patronage so it would give us a good sample size of patrons for our observation. We set up at 6:30am and observed patrons through till 8:30. We then gathered as a group to summarize our findings:

Field Test Observations


I think this is one of my favorite evaluation methods as it gives us a real observable effect on our design causing practical behavior changes, and that speaks a lot louder than any written theories will do on its own. The biggest downside I see is that such types of testing can be hard to do right. If patrons know they’re being observed, it would almost definitely change their behavior, if they don’t know they’re being observed, it goes a bit against ethics to observe without their knowledge. While as students, people tend to be more forgiving of minor intrusions and forgive minor over stepping, but may not be an option once we are in our professional lives.

Concepts Reloaded and Matrix

From the feedback and critique we received, we decided on a few points:

– We would keep the kerb-side location, but look into limitations (eg. distance from road edge for safety).

– We would keep the idea of queuing to improve flow but with a gentler approach.

– Our design would be modular to adapt to many spaces.

– Our design would consist of 3 levels, with the focus on the floor and seat level, shelter to be developed as an accompaniment rather than as a focus.


We made the decision not to focus on shelter since even designing a shelter on its own is a pretty hefty task. We felt 3 weeks would not be sufficient time for us to develop all 3 components to a high standard, and would rather execute our core ideas of seat and floor design well.


We decided that while there were good and bad points to the 3 concepts we initially presented, we also wanted to generate some concepts that directly reflect the points raised in feed back. We made a few more models and came up with 3 more concepts, and put all our concepts into a matrix.



What was the most surprising to me, was that the concepts that ranked highest was the floor concept for our 1st set of concepts, and a seat concept from our feedback concepts. This was surprising to me, because I had expected our newer concepts to rank better in the separate criteria since a lot of our criteria came from our feedback. What we found instead was that them most simplistic concepts actually did the best, simply because what they did they really did well, whereas the others tended to have even scores across a number of areas.

Seat concept483959_10200909289643290_239026652_n


I think what I found to be most useful this time was the matrix, since it gave us a surprising result, we went back and looked at those designs and found that we had under estimated their value a bit just because they were simple. I think the matrix provides a more objective way of evaluating ideas so we don’t over or undervalue based simply on intuition (can be a common pitfall for designers)