City of Ann Arbor Compost Optimization
Project Format: Semester-long group client project for SI 538: Citizen Interaction Design
Role: UX Researcher/UX Designer
Time Span: September 2021 - December 2021 (4 months)
Task: Optimize Ann Arbor residents’ compost drop-off experience
Frustration-filled Compost Drop-Off
I worked with a group of three other graduate students and the City of Ann Arbor Public Works office to analyze and optimize the city’s compost drop-off process.
Our partner, the City of Ann Arbor’s Public Works department, requested help streamlining their compost drop-off process. The process was supposed to be completely automated with the purchase and integration of a public-facing kiosk on the scale, but usability problems at the compost scale created a large influx of requests asking for assistance with the process from the scale call button. These help requests are monitored and answered by the administrative assistants at the Public Works office, and make their already heavy workload more difficult.
The usability problems take place throughout the compost drop-off process, with the first occurring when residents enter the compost center from the road. They are met with the confusing sight of a road and a bridge-like structure, with no clear indication of where to proceed to drop-off or pick-up their compost. Assuming that they successfully drive onto the right side of the scale, the resident is then prompted to interact with the scale kiosk. The resident is asked to scan their license, select the material they are picking up or dropping off, and collect a ticket to show to the booth where they actually exchange the compost. The kiosk created more confusion for residents as there was a large number of labels that fought for the user’s attention. In addition, residents are asked to use the arrow and enter keys on a keypad to select their compost material, which added more complexity to the process.
Gathering Information from Relevant Stakeholders
Our first step after meeting our community partners was to do a site visit to the scale to see the process as a whole in person. We were able to see a demonstration first-hand of how a typical resident would use the kiosk as well as learn of some high level errors that residents often have. We were also able to speak with the compost site vendor WeCare Denali and learned of common issues residents had when picking up or dropping off compost, leaves, etc. After this initial meeting the team discussed potential solutions to the issues uncovered in our meetings with our community partners, the kiosk demonstration, and speaking with WeCare Denali staff but kept it high-level so as to not commit too much to a particular solution or path.
With our high level concepts in mind we had a second meeting with our community partners to learn more information about the particulars of the kiosk, the feasibility of some of the ideas we had in mind, and to get an understanding of what parts of the kiosk system as a whole Paradigm has more control of versus what parts our community partners have more control over. Following this meeting we requested our community partners keep a log of error calls they received so we could better understand where to focus our efforts.
The team then determined that in order to address any software side issues with the kiosk we would need to have a call with a representative from Paradigm. During this call we learned ways we could update the kiosk UI as well as gained a further understanding of the complexity of the address verification issue. Based on our conversation with Paradigm, coupled with our analysis of the current state of the kiosk system, we decided to focus on improving the user experience of the compost dropoff process from entering the facility to leaving the scale to drop off or pick up materials. This decision was made based on the timeline given to us within a one semester course as well as information received from call logs showing address related issues were a relatively small portion of the calls received.
Once we had an idea of what issues to solve we developed an interactive prototype of the scale as it is and conducted a series of user tests with homeowners and other users interested in composting. Based on our findings and observations we developed an updated prototype of the system to address common concerns and errors. We then further refined our prototype based on a second round of user testing feedback.
A Pre-existing System Creates Solution Limitations
When creating our solution we needed to be cognizant of a few limitations. The first limitation stemmed from the Paradigm software, as it has limitations on screen layout options, a limit on the amount of pictures on the screen, and can only display Latin script on screen.
The second limitation is that residents are not the only people that use the scale. Contractors use the scale on a daily or weekly basis, and are comfortable with the current flow of the system. We needed to be sure that any changes we make to the system did not hinder the contractors, but also help the residents.
Our third limitation was regarding address verification of residents at the scale kiosk. Residents were being incorrectly categorized as a non-resident because the address on their drivers license was not within Ann Arbor. With the time constraint of the semester, and our inexperience with databases, our solutions to this problem were limited. We also realized our residents were encountering other problems before they could even scan their license. With our residents’ needs in mind, we decided to address other, more prominent usability issues with the compost drop-off process, but did briefly address the issues through our conversation with Paradigm.
Initial Solution Prototypes
After conducting our due diligence to better understand the task at hand we created some low-fidelity prototypes that could improve the user composting experience. We created a simplified version of the kiosk to prevent information overload, a more user-friendly material selection screen for the kiosk, and also made some suggestions to improve the navigation on the compost page of. thePublic Works website.
User Testing Plan
With these basic ideas in mind, we set out to gather user feedback from residents. We quickly realized that it was very difficult to get people to conduct a user test while actually on the scale, and there were few residents using it while we were there. So, we adapted and made a virtual walkthrough of the compost drop-off process in Figma. This allowed us to conduct user tests remotely, and we had much more success. We conducted user tests on both the original process without any of our modifications, as well as on our modified prototypes of the composting process.
During these user tests, we made sure to note which areas of the process users most struggled with, as well as timed their interactions and made sure to record what stood out to the participant regarding signage and other important parts of the process.
Recommendations for Frustration-Free Composting
We proposed a few different solutions to address the problem:
Paint road lines around the scale to direct traffic to move in the correct manner. Through our user testing, we found that many users felt intimidated just driving into the facility, and giving them clear direction on where to go is important to the overall success of their visit.
Remove most of the text and directions on the front of the kiosk. The current amount of direction text on the kiosk causes users to glaze over and actually ignore even more of the directions. Simplifying the instructions on the kiosk and having the only source of truth being the monitor makes it clear that users will get all the information they need on the kiosk screen.
Add pictures and streamline the kiosk UI for residents. The current instructions on the kiosk screen are in the green at the very top and mention contractors/city vehicles before residents, changing the order of this catches the attention of residents but doesn’t affect the usability of workers/contractors. Adding pictures with the selections mapped to singular numbers also makes it easier for residents to see on the screen what they are dropping off and press a singular number to select their payload instead of having to use the arrow and enter keys to scroll and select the material.
How to keep the scale Frustration-Free
Implement
Due to current weather conditions painting lines would likely have to wait until spring to maximize how long the arrows would last. In the meantime, reaching out to Paradigm to show current wireframe mockups of an updated kiosk interface can begin now, pending approval of our suggestions. As we do not know the specifics of the Paradigm contract and what services are included in that we can not give an estimate of the cost, if any, that would be involved in updating the kiosk but our solutions were designed in a way to require no new or updated equipment costs.
Maintain
The lines would most likely need to be repainted every year to every other year because of Ann Arbor’s weather conditions and the wear and tear of salt, snow, and trucks would have on them. The city may have negotiated rates on street painting but from online research, we believe the cost would be around $400-800.
The kiosk screen with composting options may have to be altered over time or as the seasons change. This will likely require the creation of a few screens that would be updated or swapped when needed.
Sustainability
Maintain the call log to keep a record of the type and frequency of errors that residents have, this would also be a good way to measure the efficacy of the implemented changes. A monthly or end-of-season review of errors will allow the Public Works Department to keep a record of error trends in case the kiosk interface needs to be further adjusted.
How to Further Improve the Composting Experience
During our call with representatives from Paradigm, they confirmed it is possible to update the central address list without wiping out all previous changes. The representative from Paradigm indicated they were open to clarifying the process to Public Works staff. This can be accomplished quickly by setting up a meeting with the Paradigm engineers, Cara the Public Works manager, Ron the Public Works engineer, and any other interested parties. By updating the list each time a resident calls in because of address verification problems the calls will eventually stop as each resident that had an issue will then be in the system. As we do not have details of the city’s contract with Paradigm we cannot say for certain if there would be any cost to implement these suggested changes. Non-monetary costs will include time on behalf of the Public Works team to reconstruct the database system to allow for new addresses. This recommendation would still require work on behalf of the administration team to add new addresses as they are encountered. Due to inconsistencies in the address databases between the SOS of Michigan and the City of Ann Arbor, another verification system may need to be created.
One of those longer-term solutions to resident errors related to address verification could include QR codes residents can receive via an online resident registration system and/or as an attachment on Ann Arbor resident water bills. Attaching the code to water bills would not cover all residential verification issues but could serve as a solid foundation. Implementing an online address verification system would serve two functions: 1) it would allow residents to preemptively resolve potential address issues prior to them occurring and 2) while the system in the short term would be an added task for office staff it could be done as time allows and would reduce the number of calls in the long term. This solution would involve Public Works staff coordinating with both the web developers Ann Arbor contracts to update and maintain the site as well as staff in the billing department to determine the feasibility of adding a QR code to water bills or adding an online registration feature to the Ann Arbor website. A QR code system would be useful not only to the composting process but could be used by other city departments as a new means of individual resident identification. Initial meetings can take place as soon as it is deemed a potential solution by Public Works staff. As this feature would involve coordination with multiple offices within Ann Arbor we cannot provide timeline milestones for implementation. Depending on the complexity of the QR system a third party may have to manage the codes and their distribution as people move in and out of the city. Costs related to this solution are hard to estimate as many factors are left undetermined.
