Digital Product Designer
Tags: iOS Design, 
Dumpster Diving, Thematic Analysis
Tools: Balsamiq, Whimsical, XD, Illustrator
Project Space: HCI Lab @ Penn State
Team: Jordan Beck, Adam Wattis, Adi Chakravarty

Inspired by the design of Apple Maps for iOS, Dive was built as a means of synthesizing insights from an initial thematic analysis of over five thousand comments obtained from The app is intended to allow dumpster divers to search for and add information about various dumpsters on an interactive map.

Project Details
Dumpster diving involves sifting through commercial or public refuse, so as to reacquire and reuse specific, discarded items of interest. The practice can be misunderstood as an undesirable counter-culture to traditional societal norms. Although individual divers may search for and acquire items of interest, their motivations to do so are wide-ranging. For example, some are motivated by poverty and food insecurity. On the other hand, some are motivated by propagating a modern response to abundant waste. Thus, there is not a single motivation to dumpster dive. Divers often have to adapt quickly to changing circumstances (e.g. the arrival of a security guard or of other divers) and employ creative methods to retrieve what is available – the most sought after items are edibles followed by clothes, electronics, and furniture.

Although there have been research studies of dumpster diving, its constituents can be resistant to the presence of outsiders who only wish to understand, rather than participate in, the practice and its political project. However, in many cases, divers make use of publicly accessible websites, wikis, subreddits, and other digital tools, which can serve as sites for doing research into the practice.

For instance, Dumpstermap is a web-based platform that enables a global community of divers to access, and contribute to, a crowdsourced index of dumpsters. Users can navigate a world map to a location of their choice to search for information about existing dumpsters or add new ones.

We performed a thematic analysis on 5,423 comments obtained from Dumpstermap (with the majority of users located in Western Europe and North America) with the goal of identifying categories of useful information. Using an inductive approach, we were able to derive inferences without a pre-existing coding framework, thereby coalescing our research processes into a data-driven thematic analysis. We generated three categories of useful information for dumpster divers: location, quality of items, and accessibility.

Utilizing the themes generated from our analysis, we designed a high fidelity prototype, Dive, in which we synthesized our research findings so as to engineer a potential solution that could aid dumpster divers in performing their activities in a more efficient and optimal manner.
My Role
This research project was conducted by a team of three members which included a postdoctoral facilitator, a developer-partner, and myself. Although we conducted the thematic analysis and came up with the process flow together, my individual contributions included conceptualizing concept sketches, wireframes, UI design assets, and the final prototype.
Thematic Analysis
The dataset obtained from, citing several accounts of experiences by dumpster divers, consisted of over 5000 comments on dumpster locations from around the world. Some comments were observed in German, Polish, French, and Swedish, although most comments were posted in English. All comments other than the ones written in English, French, and Swedish (the languages we were capable of interpreting) were discarded. A thematic analysis was conducted on the dataset in order to gain appropriate insight into what dumpster divers considered to be important information that could be potentially conveyed to their fellow divers. The interest was to develop an essence of what concerned them in their activities.

We conducted our analysis based on the steps as outlined by Braun and Clarke, this approach was employed so as to extrapolate a deeper meaning from the data, with the themes strongly linked to the data itself.

The dataset was initially skimmed over to gain a high-view picture of the format pertinent to each comment. This format was noted to be similar in most cases; some contained addresses or descriptions of dumpster locations. Most comments explained what was found and the condition of the items. In general, a comment described a single account of a dumpster diver’s experience when visiting a dumpster.

Dumpster behind Sunterra. The dumpster is located in the loading dock, and is an elevated dumpster. Good to be careful here, as you have to climb up a bit to look in. Also be careful going in, as the dumpster is deep. Usually not much in here, though when they throw out bread it is great quality, packaged, and in abundance.

In the pursuit of finding out what dumpster divers cared about and to generate codes, we jotted down any mention of a word, idea, or concept that kept frequently occurring while reading the comments. Some codes were easily identifiable. Others took more time to discover. Some of the most striking codes generated were:

  • Directions to a particular dumpster.
  • Specific descriptions of items found.
  • Obstacles encountered when visiting a dumpster.
  • Simple tips for dumpsters that may be left unchecked.

Through our analysis, we identified three main themes, namely location, quality of items, and accessibility.

Location, and words of synonymous meaning and relevance, appeared in several instances within the dataset. Typical mentions included directions to a dumpster at some obscure location. Other comments included addresses, or in many cases store names to describe a dumpster. This category was important enough to be a theme in of itself.

Trader Joe’s: Really solid dumpster where friends of Nuclearphyllis scored nearly 30 bottles of wine once. Best one around for you know what!

Huge piles of bags on blindingly lit stretch of 2nd avenuenot for the shy. Similar to what’s found at Gristede’s in other parts of town.

With respect to our second theme, we had initially derived codes that suggested quality of food as a potential theme, since many of the comments did highlight the discovery of various edible foods. Although, after closer examination of the comments, it became apparent that artifacts other than food could have also been found while dumpster diving. In order to fit this condition under the same theme, it was changed to quality of items instead. This theme amply fit the array of foods, as well as clothing and miscellaneous items that a diver might find. However, it centers itself around the fact that the state of the items, or quality, also mattered to the divers.

Supermarket bins. Flowers, newspapers and sometimes subs (half eaten subs and bread).

This health food store has had small but good finds recently including produce, juice and packaged salads.

Our final theme, accessibility, came to be after realizing the many obstacles that a diver could potentially encounter in the pursuit of reacquiring items from a given dumpster. While assessing the comments found in the dataset, physical accessibility was observed as the most obvious of the challenges. These generally included gates, fences or locks. There were other, fewer occurring mentions of obstacles we found as well, such as watchdogs and security guards. Accessibility was therefore identified as an important aspect a diver needed to take into consideration and also to communicate to others.

Hemköp: Security guards tend to check the area at 23:00 so beware!

Access via holes in the fence. No need for climbing.

Through our analysis, we were able to develop a broad understanding of the multitude of factors affecting divers alongside deeper insights into how such an activity is conducted by different members. We realized that the use of technology amongst participants to enhance one’s experience was an important subset of the activity.
Process Flow
Following the above analysis, our next step was to translate each theme into a feature set using which we could come up with a high-level process flow.

Considering the ubiquitous nature of accurate location discovery and its key relevance to dumpster divers, we reflected upon this theme through the following feature set:

  • Adding a dumpster by sharing one’s current location, dropping a location-based pin, or searching for a particular location.
  • Accessing nearby search recommendations based on location proximity.
  • Viewing an address specific to a dumpster listing in a given location.

Given the high importance of assessing the quality of items found at a dumpster, this theme represented central features that were translated for Dive:

  • Adding a dumpster by defining the rating of items found, any additional user comments, and a supplementary image uploaded alongside to support a user’s findings.
  • Contributing to a dumpster via an identical process flow as described above.
  • Filtering through items available nearby under each category–food, clothing, or miscellaneous, on the basis of proximity.
  • Viewing a specific user-defined rating in an uploaded post.

Divers often operate in a gray area where there is a risk of running into obstacles of inaccessibility. These findings, relating to such accessibility risks, were incorporated into Dive in the form of features, as mentioned below:

  • Specifying risks to accessibility when adding a dumpster. 
  • Reestablishing accessibility risks when contributing to a dumpster. 
  • Viewing useful, glanceable information relating to accessibility within a particular dumpster listing.

Based on the above feature set, we came up with the following process flow for Dive:

We sought to utilize this process flow to further our understanding and created low-fidelity concept sketches, mid-fidelity wireframes, and a high-fidelity prototype intended to aid divers by fostering a positive influence upon their experiences.
Low-Fidelity Concept Sketches
My first step after we came up with the process flow was to design a low-fidelity concept sketch using Balsamiq showcasing a prospective user's journey in the app. Although we did settle on using Apple Maps for iOS as the preliminary template for Dive, it was crucial for me to get a high-level sense of how the app would eventually function.

Mid-Fidelity Wireframes
My next step was to translate the low-fidelity concept sketch into a mid-fidelity wireframe using Whimsical. This step helped me in clarifying the visual design for the app and aid in the design of UI assets.

High-Fidelity Prototype
The themes were used to justify the basis for a set of features that helped in the creation of a prototype platform for dumpster divers, known as Dive. This prototype was built using Adobe XD. I utilized the features to create a coherent interface that was closely mapped to one or more themes from the analysis.

Adding a Dumpster
As previously expressed, users can add a dumpster to a particular location based on the identification of items, rating the quality of found items, and specifying potential risks to accessibility.

Nearby Listings and Availability Filters
Users can find dumpsters near their location or search for a dumpster by place or address. Additionally, they can filter dumpsters around them and in other areas based on the types of items previously found by other divers.

Dumpster Information and Interactions
Users can view information specific to a dumpster such as the address, categories of available items, and posts contributed by other divers. One can also contribute to a dumpster using a process similar to adding a dumpster. Other interactions include getting directions to a particular dumpster, viewing an individual post, and marking a dumpster.

Even though we adopted a data-driven approach to our analysis and subsequent design ideology, there are many dimensions along which a user’s needs and experiences may vary. An interface may be based upon the principles of realism or abstraction. Such choices ultimately dictate the usability of a particular design and could accommodate for several implementations and permutations of user experiences. Our ideology strived to incorporate two fundamental tenets of interaction design, specifically under the guidelines set forth to propagate direct manipulation:

  • Showing users immediately how their actions are furthering their goals.
  • Helping users gain mastery and feel in control.

Since our design motivations for Dive’s current implementation are quite straightforward and ideologically fixed, there is always the potential for other implementations to be developed, irrespective of the approach at hand. Furthermore, considering that our designs were derived on the basis of the themes we identified, it may be almost certain that a thematic analysis on the same dataset performed by other researchers would likely yield very different codes, patterns, and hence, lead to a very different prototype being created.
Dumpster diving is a practice that has been reported to take place in various parts around the world. It is typically more than a response to food insecurity, at times being a reaction to conspicuous consumption. This niche community has been largely unresearched, likely because dumpster divers tend to operate with a certain belief that they don’t want any unwanted attention and are therefore more difficult to reach out to and communicate with. However, despite being a widely practiced yet undertheorized activity, there is a lively community that exists online on different forums and websites that facilitate divers in their pursuit of discarded items.

By means of executing a thematic analysis on several thousand comments, from various dumpster divers across North America and Western Europe, we successfully developed a prototype for a mobile application. Our primary goal with regards to such an approach was to demonstrate the viability of our prototype in not only fulfilling but also influencing a diver’s core experience when indulging in the practice. These motivations, while not inherently critical of other approaches, were driven by the potentiality for supporting the diverse experiences of a self-regulated community.

👉 Check out the clickable prototype at
👉

© 2021.