Intelligent auto store
UX/UI: Machine learning in retail stores.
The power of machine learning makes it a useful tool in just about every industry. For this project we were tasked to incorporate machine learning and facial recognition in an retail environment that sells car parts. The final product was a community based help board that lets a user know if someone in the store is an expert, an achiever, or a learner. Collaborating on this project was Victoria Gerson, Ann Frohbose, and Ashlet Anderson.
Learner : Limited knowledge of cars. Has someone else do it.
Achiever : Works on their car own. Interested in learning more about cars.
Expert : Enjoys working on not just their own car but others as well.
To start this exploration we needed to understand how users, in this case an expert, perceived machine learning. We started by polling people who frequent auto stores and constructed a hopes and fears matrix to establish the users comfort level towards machine learning and privacy.
hopes and fears
STORE OBSERVATIONS & INTERVIEWS
We went to local stores to understand how our users interact with the store by doing fly on the wall observations using observation sheets to document certain interactions. After users use the space and what potential pain points we can leverage machine learning to solve. We paired up these observations with interviews of store managers to understand how the store procedures worked from day to day.
"as is" journey MAP
"to be" journey MAP
Understanding the users pain points and their current experience, we constructed an "as is" journey map that showed the users current path throughout a visit to the automotive store. Then we placed the technology into the journey and formed a "to be" journey map to show how these pain points will be alleviated by the intended design.
TECHNICAL LIMITATION FEEDBACK
We presented our persona's, journey maps and findings from the hopes and fears matrix to data scientists to understand our technical limitations The feedback gave us a clear understanding on the limits of machine learning and our intended design solutions.
Early on we discovered a few things:
Experts had a passion for helping people
Learners and achievers want to learn from experts.
Employees at the store are sometimes busy and need assistance.
People who live around the store tend to be customers for longer periods.
This shifted our perspective of the store towards being an ecosystem and something that our designs can facilitate for longer periods of time. To explore the give and take relationship between the expert and the other two persons we used a screen in the store as a communication hub to connect people who frequent the store. Using facial recognition the store can track how often a person comes into the store and recommend certain connections.
The screen we used was located in the center of the store and was 5 feet tall and 2 feet wide. When designing the experience we wanted it to be accessible, both for color blind and for people in wheelchairs, and engaging to catch the attention of people walking throughout the store.
The lofi iterations explored a network approach to give learners and achievers the information needed to find an expert and know what their limitations were for helping. We looked at how ranking systems and machine learning can use purchase records and store frequency to update experts profiles without user interaction.
We invited users to test our experiences on both digital screens and life size printed screens. The printed screens allowed us to understand hierarchy and where users look when they first approach the screen. Users also pointed out how users could take tests to confirm their skill level on certain automotive topics. We took that feedback and offered an expert the opportunity to earn rewards for helping other customers within our design system and breaking their skill level down into sections of specialty.
HIFI PROTOTYPES and Scenario Video
To conclude this project we created hifi screen mockups and shot a scenario video within the store. The final scenario video and process work was presented to the client in their headquarters in downtown Raleigh.
Thank you to our actors : Adam Morgan (Luis) and Hal Meeks (Craig)