Photo capture
User takes requested photos with the phone app.
ACV Auctions / Guided Photo Capture
Guided photo capture for smart damage detection.
Redesigning guided vehicle photography so customers can capture clearer evidence for damage detection and pricing.
- Role
- Product Designer
- Focus
- Research and mobile UX/UI improvements
- Platform
- Mobile app
- Scope
- Improve experience from UXR insights
- Timeframe
- 4 weeks
- Delivery
- Demo and SDK; implemented in dealerships by two major automotive brands.
Project summary
Vehicle photos are the starting point for a condition report.
The app lets users photograph their vehicle to receive a condition report and an estimated selling price. AI analyzes the photos for exterior damage and estimates potential repair costs.
Condition analysis process
AI validation
Photos are reviewed for good quality and details with AI.
AI damage detection
If the photos meet the quality criteria, the app identifies damage on the vehicle.
Report creation
A condition report with identified damages is created for price estimation.
UX research findings
Four capture problems
- Overlays did not fit different vehicle shapes.
- Thin guiding lines were hard to see in sunlight.
- Instructions did not explain where to stand or how high to hold the phone.
- Automotive terms and left/right directions were confusing.
Discovered opportunity
Make a bad photo recoverable.
Make the capture state clear, then let people review or retake a photo without losing their place. Prior research also found that a processing spinner made some consumers think they had to wait before continuing.
Problem statement
How can we make it easy for anyone to compose high-quality vehicle photos?
My approach
I translated the research findings into a capture experience built around familiar mobile camera and gallery patterns. I focused on four priorities: showing users where to stand and how to move, simplifying overlays and labels, making photo feedback clear, and keeping review and retakes within the capture flow.
Guide positioning and movement
Show users where to stand, how to hold the phone, and how to move between shots.
Simplify overlays and labels
Simplify the silhouette, strengthen contrast, and replace specialist terms.
Clarify photo feedback
Explain photo quality issues so users can understand what needs to change.
Keep review and retakes connected
Keep photo status, review, and retakes within the capture flow.
Exploring how to guide the shot
Bounding box
Geometric wireframes provide a target box to help users frame the vehicle.
Smart overlay
The overlay turns green when the vehicle is aligned with the expected framing, and red while it remains misaligned.
Position and phone height
Instructions and icons show where to stand around the vehicle and how high to hold the phone.
Transition animation
An animation demonstrates how to move from one photo position to the next.
Some concepts were not clear or simple enough; others required too much development time or faced technical constraints. What we learned from these explorations shaped the improvements that follow.
Improvements
Simplified overlays
Stronger visibility
Use thicker lines and greater color contrast.
Fewer details
Remove details that do not help compose the required view.
More adaptable silhouettes
Use a simpler vehicle shape that remains useful across body styles.
Instructions
Persistent instructions
Keep the starting point and next movement visible during capture.
On-demand instructions
Provide a more detailed position cue on demand, aligned with the scene animation for that shot.
Scene animations
Animations show users where to stand and how to frame each shot.
Capture navigation
Flexible photo sequence
A sequential path can guide consumers around the car and reduce backtracking while keeping navigation freedom through the bottom controls.
Status in context
Status and photo details are kept within the camera interface to support moving between capture and review in non-sequential flows.
Review and submission.
The gallery brings captured and remaining shots into one view, with a route back to the camera or a photo detail for reviews and retakes.
Outcome
What user testing showed
- Photo instructions were clearer.
- The interface felt more comfortable.
- Rejection feedback helped users understand the problem.
- Rejected photos were easier to retake.
What still needed work
Navigation remained unclear and needed further refinement. These findings were qualitative; we did not measure changes in task time or retake rates.
Expected impact
Clearer guidance and easier recovery were intended to improve photo quality and reduce retakes. We did not measure those outcomes.
Implemented in dealerships
The work resulted in a demo and an SDK that two major automotive brands implemented in their dealerships.
Takeaway
More guidance and support on demand.
The interface should make each shot clear on its own, with additional guidance on positioning, framing, and retakes available when users need it.