Vehicle Inspection App
Self-service mobile app for returning leased vehicles, powered by machine learning
The lease-end process is long, messy, and full of manual reviews and paperwork.
Critical to this process is a damage assessment before vehicle return. Currently, this is a time-consuming process that leaves borrowers with no idea how much they may owe when they return their vehicle until the inspection process is complete.
I designed a solution to enable customers to create a comprehensive report of vehicle return condition without third-party interaction and provide lenders with a way to eliminate manual processes & paperwork.
For customers, this reduces a big pain point in the lease return process. Plus, they can take immediate action on their vehicle condition report independently, take advantage of offers, and save money.
With knowledge of repair costs & follow-on margins, lenders can provide offers to fix the damage for competitive rates in exchange for lease renewal.
My Role
- Product strategy and ideation
- User research and testing
- Wireframing and iteration
- Low and high fidelity prototyping
- Interaction design
Product strategy and ideation
I conducted workshops and interviews with subject matter experts, created design concepts, and mapped process flows to understand business challenges and validate the problem statement.
Prototyping
After identifying the inspection as a critical part of the lease return process, I developed prototypes to test the feasibility of self-service for vehicle inspection.
The prototypes focused both on user interaction as well as the current technology capabilities in identifying damage. I worked with an engineer to implement cloud image processing capabilities in a functional prototype to process images in real-time.
User Testing
Once a solid foundation was built, I conducted user tests to understand the burden of completing a self-inspection from a customer perspective.
User Tests was incredibly fun and enlightening to see how customers would interact with the app and their vehicle in real-world scenarios.
There were many immediate learnings, including understanding about expected sequencing for photo taking in and around the vehicle.
MVP Design
Incorporating feedback from target users, I iteratively designed an MVP built as a mobile web app for conducting self-inspections.
Machine Learning - Admin User Interface
As part of the release roadmap, the machine learning photo processing would need to be trained. Working with an engineer, I designed an admin interface that addressed training and validating user inputs from both the customer and the admin to power the machine learning model.
At the end of testing and MVP build out, the app experience provides customers a quick and easy way to get a damage report for the vehicle.
