Auto Park Assist Re-Design
One of my first big projects at GM was an overhaul of the Auto Park Assist user experience, showcased in the Cadillac Lyriq. It was a great challenge because I was only able to change the interface – not how the feature functioned behind the scenes.
Understanding the Barriers
I collaborated with a UX researcher to gain a deeper understanding of the issues users faced with Auto Park Assist. To determine why users felt the feature was not ‘working’, we first tested it ourselves in a vehicle.
We then conducted some exploratory user research to better understand our participants’ mental model of how the feature ‘should’ work. To achieve this, we described the feature to them and asked them to demonstrate and explain how they thought it would function, pushing a Hot Wheels car across a mat featuring a printed parking lot.
Old implementation of GM’s Auto Park Assist feature.
Aggregating the Data
From our recorded sessions, I collected the main expectations and pain points of our participants and compiled them using an affinity wall and discovered three main issues:
The users' conceptual understanding (i.e., mental model) of how the feature operated did not align with its actual engineering implementation
Users were nervous because they weren’t sure what the vehicle could ‘see,’ or what it was planning to do next
Users wanted to feel in control of the vehicle, even when it was in an automatic state
Iteration & Testing
Using our collected data , I iterated and performed observational user testing to ensure my designs improved the experience.
To address disparities in our users’ mental model, I added clear explanations and visuals at each stage, employing a friendly, conversational tone for better comprehension and trust.
Sometimes automatic actions can create a negative experience for users, especially when the technology is unfamiliar or the automatic action happens unexpectedly. By increasing communication between the vehicle and driver, we ensured the vehicle was always helping the user feel in control by providing clear choices for the user and always asking for permission before taking an action.
Due to the lack of clarity when in automatic mode, I revamped the language and visual progress indicators to clearly and consistently explain what it was doing in real-time.
Final Experience
By the end of this process, I designed a much more intuitive experience that tested exceedingly well with our participants and is now found in GM vehicles with Auto Park Assist today.