Background and Context

As a group, we brainstormed a bunch of ideas within the prescribed theme of "quantified campus". Inspired by our own experiences as new students on campus, we decided to research physical navigation on-campus as our project's problem space.

Goal

To design a solution that would enable navigation and way finding on Georgia Tech's campus.

User Group

Students at Georgia Tech

Process

Process

User Research - Phase 1

We conducted a preliminary literature review to understand the domain of wayfinding and navigation better.

We hosted an online survey as a quick and dirty approach , to understand current approaches people use to navigate and the gaps in them and opportunities for improvement there of.

In order to understand our problem space better, we decided to talk to a few representatives from our target group as well as people on Gatech campus, who can be thought of as information sources for our target group.

We used the notes from our interview data, and performed an affinity mapping

User Research - Phase 2

We designed an activity for our users to take part in.With some probes, we asked them to paint us a picture of how campus looks like in their heads , thinking out loud as they constructed the layout. We did this activity in order to get a deep dive into understanding how users build and use mental models, to navigate the campus. The Following image shows a layout constructed by one of our users.

A total of 4 users participated, 2 of them had spent 3+ years on campus and 2 of them were students in their first semester on campus All sessions were video recorded , and later analysed.
The notes from these sessions and the videos, were then mapped into an affinity diagram, as seen below.

Research insights

Following are some of the major insights that we were able to gather at the end of our research process

  1. There is a need for a spatial sense of the campus.
  2. There is a selective need to know places.
  3. Building navigation is a contextual even to people who are not new to Georgia Tech.
  4. Users have more visual recall ability on routes they have walked/walk regularly.
  5. Users may have central landmark for reference.
  6. Users have their own spatial metaphors and associations.
  7. Construction zones interfere with navigation.
  8. Users associate events/food/crowdedness etc. with certain places.
  9. If/When lost while navigating to an unfamiliar location on campus, users tend to seek the help of friends.

Design process

Taking the research insights forward, as a group we brainstormed for multiple ideas. To spark the process of generating design ideas, we conducted a round of quick brainstorming, with each member of the group contributing 5-7 ideas
Through the process of aggregating our ideas, we discovered that there was a lot of similarity as well as compatibility between our ideas, so we classified them by scope
(i.e. whether the idea constituted a standalone solution or a feature that could be tacked on to another solution).
Majority of our ideas fell into three broad “buckets” of solution space : 1.Leveraging Wearble devices 2.Using Apps and kiosks and 3. Crowdsourcing data and Interactive environments.
We then created 3 physical low-fidelity prototypes that represented our 3 top broad ideas and sought feedback from peers.

Initial Design idea #1
Initial Design idea #2
Initial Design idea #3

Based on the feedback that we received, we decided to combine the strengths of the individual design ideas , in the form of a mobile app. The idea was to let the user customize their own map, and navigate the campus through his/her familiar buildings. This map-like app also combines crowdsourced information about events and activities around campus, and lets the user share his/her model of campus including buildings and paths with the other users of the app.

The prototype

Building on the main ideas of the customizable maps and the social navigation concepts, we built a medium fidelity prototype to allow users to build a familiar model of the campus using landmarks and their routine paths. Based on these landmarks known or recommended to the user, the app will (A) display routes shared by others on the platform closest to these landmarks, (B) use the model to ask for directions from others who know better, and (C) populate the maps for events and related paths reported by the larger community or key personnel.

Evaluation - Cognitive Walkthrough

We decided to conduct an initial round of evaluation using the cognitive walkthrough approach to review if a) the task flows were adequate to convey the core tasks we had identified and b) understand the expectations from the prototype so we could set user expectation as we intended to use our limited medium-fidelity prototype after their inputs to elicit feedback from usability testing.

What we learnt

The main implications of the findings from this evaluation activity was that the core idea of the application - the campus map model - as an abstract concept could not be easily understood without explanation.The participants pointed out that once the map model concept is learnt, the system felt natural. We understood that the sections of the applications needed to be set in a more natural way for the user to recall and understand within specific contexts. This is especially true for new users to explore applicable options before they have even a moderately formed mental model of campus.

Another thing we observed was that although the top view is the commonly used view for navigation systems, we detected it may not be the more natural visualization strategy for all the users, in particular when the user was trying to identify places in an area the user was not acquainted with yet. In this sense, including a picture mode or 3D mode to select landmarks and to navigate seemd to be more aligned to the users needs.

The following is our final high fidelity prototype.

Final prototype

More Thoughts

This project was particularly interesting and stands out to me because of the nature of the problem. When we initially set out to take this problem on, what we heard often was "how is this any different from existing map solutions". It was such a challenge to understand how our target users made sense of this limited and constant physical space, and how that model changed over time. When we taped into the mental models, a lot of the features seemed so obvious and the entire project received good feedback.

If we had more time, we would have also considered including bus/cycling routes and also considered integration with Georgia Tech's CMS websites for events and locations.

When and Where
August 2016 - Dec 2016,
Course project @ GT - Fundamentals of HCI
Team Members
Research Methods
Prototyping tools
Deliverables
High Fidelity Prototype
Project Report