Ford Expands AI Vision Checks to Catch Assembly Defects Earlier
Ford is deploying AiTriz and MAIVS computer-vision systems to detect assembly defects in real time. The checks can help workers fix problems earlier, potentially avoiding costly rework and reducing quality risks. Ford’s next challenge is demonstrating that broader use of the systems measurably lowers warranty and recall costs.
USAGEFUTURETOOLSWORK
The AI Maker
10/15/20262 min read
Ford is using AI-powered cameras on factory floors to spot assembly defects while vehicles are still moving down the line. The systems, AiTriz and MAIVS, check whether components are correctly installed, helping workers address problems before they require extensive repairs or lead to customer-facing failures.
The tools target a difficult quality-control problem at the Dearborn Truck Plant, where more than 300,000 F-150 pickups are assembled each year across a complex range of trims, wiring, electrical hardware and other configurations. A connector that appears attached may not be fully seated, and the difference can be difficult to detect by touch or sight amid factory noise and protective gloves.
Ford’s two systems use different approaches. MAIVS, introduced in January 2024, analyzes still images captured by smartphones mounted on 3D-printed stands. AiTriz, implemented in December 2024, uses machine learning and live video to identify millimeter-scale misalignments. Ford engineer Beatriz Garcia Collado, who developed AiTriz in Spain, is the source of the system’s name.
MAIVS is now installed at nearly 700 stations, while AiTriz operates at 35 stations across North America. Still-image checks can confirm that expected parts are present, while live video can provide more adaptable inspection when a view is briefly blocked by a worker or vehicle component.
The systems are designed to identify faults where they occur, rather than waiting for end-of-line inspections or later checkpoints. Ford manufacturing staff say earlier detection can prevent repairs that might otherwise require removing seats or carpets to reach an electrical connection. As vehicles add screens, sensors and computing hardware, a loose connection can create problems that are harder to diagnose after assembly.
The business case is significant for an automaker facing substantial recall and warranty costs. The reporting described Ford as having recorded 94 recalls by early August, more than any other major car brand at that point. A separate fuel-leak recall affecting 694,271 Bronco Sport and Escape vehicles was estimated by the Detroit Free Press to cost $570 million.
Ford’s factory teams say the vision systems improve in-station process control, but their impact on recalls has not yet been established. Morningstar analyst David Whiston described effective AI as a potential way to reduce recalls over the long term, while cautioning that results are not guaranteed. Preventive checks in manufacturing, design and engineering could all contribute, but the systems alone cannot address every source of defects.
Ford says the tools are intended to support plant operators, not replace them. For manufacturers, the deployment offers a practical example of computer vision being applied to repetitive, high-volume inspections: AI flags subtle deviations, while workers remain responsible for production and corrective action. Whether the approach reduces warranty costs at scale will depend on measured results as Ford extends the technology to more stations and increasingly complex vehicles.
Your Data, Your Insights
Unlock the power of your data effortlessly. Update it continuously. Automatically.
Answers
Sign up NOW
info at aimaker.com
© 2024. All rights reserved. Terms and Conditions | Privacy Policy