NCR Accessibility Kiosk

Overview

Problem Focus

NCR is one of the largest kiosk providers, selling self-service kiosks, ATM machines, and more in 180+ countries. The current self-service NCR kiosks need an upgrade to the accessibility features for users that are visually impaired. While NCR’s kiosks contain several accessibility features for visually impaired users, there is still more progress to be made in this area. In a semester-long collaboration, my team created a self-checkout grocery store kiosk concept that addresses visually-impaired user pain points and aligns with NCR’s business goals.

Solution

Contribution

As the lead UX researcher, I designed and created our interview protocol and survey, planned our research timeline, recruited participants and conducted interviews, and led expert and usability testing. I additionally aided in the creation of concept sketches and wireframes for our divergent design phase, and I designed any voice assistant technology in the prototype.

I served as the project leader for the end of our design phase in this time I defined the meeting agendas, prioritized tasks by identifying team strengths and supported team members, led discussions, and served as a point of contact for our NCR stakeholders.

The Team

Research Process

Generative Research

Guiding Questions

To understand visually impaired users better we conducted multiple research activities. A few questions that we wanted to answer with our activities:

1. How does NCR's current kiosk system work?

2. How do our target audience utilize assistive technologies? Is there a common preference for a feature?

3. How do our users use current self-checkout kiosks? What are their challenges and what goes well?

Competitive Analysis

We wanted to understand what accessibility feature were already available to users in self-service kiosks that were made by competitors (Diebold Nixford, Posiflex, Toshiba, Emboss). We looked at ADA Compliance, support for Assistive Technology products, customizability of the kiosk screen, and ease of use.

Findings

We found that a disparity existed among top kiosk companies when implementing accessibility features, and overall adding such features may not be a high priority.

Semi-structured Interview (n = 3)

Survey: Visually-Impaired People and Support Network (n = 37)

I designed and deployed a survey using Qualtrics because of its high accessibility to visually impaired users and their support network. The survey helped us reach a wider audience and supplement the qualitative data we obtained from interviews, competitor analysis, and observations.

We wanted to find out users' likes and dislikes in regards to self-service kiosks in grocery stores, their preferences in regards to assistive technologies, and we also wanted to understand how people who work with blind individuals or family members of people with visual impairments may help them.

Findings

Screen readers, screen magnification, and high contrast mode are the most common assistive technologies among respondents.


64% of visually-impaired respondents get groceries on their own at least once a week . They do not necessarily rely on others for help while shopping

Participants preferred scanning items at kiosks, and want a large screen size and adjustable text settings.

Synthesis: Affinity Mapping

Generating insights together was a very rewarding yet perplexing part of our research process. We sifted through the insights from our generative research as a team and organized them in an affinity map. This helped us identify the top pain points our users encountered and create design implications for our design.

Top Pain Points

From the research conducted, we prioritzed several challenges our target audience had using a self-service kiosk. Below are the major pain points experienced by the users that were prioritized for the first iterations. Priority was chosen based on common themes expressed through generative research and feedback from users.

Design Implications

From the pain points, as a team we brainstormed several top needs our design would address to focus our efforts and ensure our design would address these pain points.

Design Process

Brainstorming

After our initial research activities, we conducted an informed brainstorming session as a team using the pain points and design implications we identified. We timed ourselves and generated ideas freely on sticky notes, and organized the ideas in a Creativity vs. Feasibility matrix to prioritze ideas that were both creative yet realistic in our timeline.

Design Ideation and Feedback

After considering the feasibility and creativity of each idea, we narrowed our ideas down to three different potential solutions that could solve our research problem. As we designed the features, we made sure to refer back to our research questions and implications to make sure the design focused on user needs.

To evaluate our design directions, we created wireframes and I moderated a usability test with two legally blind users. We identified sample tasks for them to perform with each wireframe and asked them follow-up questions about their experience.

Below are the three design ideas as well as some feedback we received from our users.

Final Solution

We considered the feedback we received from our testing session and decided on an idea that would integrate elements from each divergent design.

We primarily went with Design 1: Tap to Change Accessibility to allow users to customize and pair their accessibility settings to the NCR kiosk, as we felt it best addressed the underlying accessibility issues with the original kiosk.

The other features include:

A 3D map to help users find items in a selected store and understand a store’s layout

Creating shopping lists prior to shopping and viewing these items on the 3D Map of their chosen grocery store.

Tap their phone to pair the app to the kiosk and send their customized accessibility settings and shopping list

If users chose to checkout at the kiosk, computer vision would be used to recognize items when they are held up in front of the kiosk in a new scanning interaction that eliminates barcodes.

Our designers created multiple iterations, and tested our idea with our users and experts to refine our design.


Low-Fidelity - Kiosk

Low Fidelity - Phone Screens

Evaluative Research

User Testing - Low Fidelity

For the low-fidelity iteration, I led user testing and feedback sessions to help us understand how to improve the usability of our solution concerning universal design. We had two feedback sessions: Experts and Moderated Usability Testing.

View Script

Findings

Likes:

Our participants thought the system seemed intuitive and clean.

They loved the feature that lets users hold their items in front of the kiosk in order to scan, eliminating the need for barcodes. 

The users also responded positively to the new interaction of pairing phones to the kiosk.

Opportunities for Improvement:

The concept of pairing the phone was initially confusing to users, and they wished they had more prompts from the kiosk and phone to guide them through the pairing process.

Adding items to their cart with the plus button was not intuitive, and clicking on the item image was preferred.

Final Iteration

User Testing - Final Prototype

To evaluate our final design, I led user testing with low to no vision users (both real or simulated with visual impairment simulation glasses) and both Accessibility and UX experts. I developed the interview script to include sample tasks and a Wizard of Oz voiceover that would simulate the voice feedback the kiosk would give a user.

View Tasks and Script

Findings

AR scanning feature, item recognition, and auditory feedback on the kiosk were overall well-received and thought to be helpful.

Pairing between phone and kiosk is a novel interaction. Users felt confused when they used the function for the first time, but they found it intriguing and easy to use after some familiarity.  

Some users had trouble seeing some information and tapping the correct buttons.

Users felt that the amount of auditory feedback during the pairing process could be increased as it would increase users’ confidence with their actions using the system.

Reflection

I came out of this process learning so much about accessibility, industry UX Research, and collaboration with industry stakeholders.

1. This process shattered any assumptions I had about visual impairments and taught me that it can be challenging to start researching for accessibility; however, the benefits and personal learning opportunities are worth it.

2. I learned the importance of ensuring designs decision are grounded in research findings. We had to continuously rereference affinity map/findings in the design stage because we had trouble prioritizing features and we added some features without solid evidence showing they were very important for our users.

3. Accessibility ultimately starts with a cleaner information architecture.

If I had more time, I would conduct observation sessions of visually-impaired individuals using kiosks to understand their process and observe any pain-points in real time. I would also have more in-person testing sessions with visually impaired individuals (and focus on recruiting earlier).