Reimagining vehicle inspections with embedded AI to help drivers complete accurate and compliant inspections faster
Company
Focus
Project Type
My Role
Year
Context & Scope
Fleet inspections are frequent, mandatory, and often completed in challenging environments. Drivers need to move quickly, managers need reliable data, and compliance requirements leave little room for error. This project explored how AI could reduce manual effort and improve inspection quality without disrupting existing driver workflows or introducing unfamiliar interaction patterns Scope included: • AI-assisted photo analysis during inspections • Guided, sequential inspection workflows • Support for both routine inspections and specific scenarios (eg. post-collision)
By using AI to quietly identify, classify, and pre-fill defect data from photos, drivers could complete inspections faster, more accurately, and with greater confidence, while maintaining compliance and data quality for fleet managers.
Problem
Manual inspections are time-consuming and error-prone. Drivers often: • Miss required photos or checklist items • Enter inconsistent defect descriptions • Rush through inspections to get back on the road Early prototypes leaned heavily on conversational prompts that increased cognitive load rather than reducing it.
Early Concept
Design Constraints
• Must remain accessible for non-technical drivers • Must work in noisy, low-attention, real-world environments • Must align with existing DVIR compliance structures • Must scale across vehicles, trailers, and equipment
Design Approach
• Making AI “quietly smart” • Seamless photo integration • Supporting real-world interruptions • Reinforcing trust and transparency • Prevent incomplete or “gamed” inspections • Post-collision detection
Mapping Photo Detection to Inspection Checklist