Industry

Automotive Safety

Client

Autoliv

DC Wheel Inspect: AI Inspection for Safety-Critical Steering Wheel Quality

Assemble steering wheel

Automating defect detection in die-cast steering wheels to improve consistency and safety

Steering wheels are safety-critical components, and defects introduced during die-casting can directly affect structural integrity. Thermal gradients, material flow, and process variation may create localised defects such as voids, cold shuts, or subtle surface anomalies that are difficult to identify reliably through manual inspection alone. In high-volume production environments, human inspection is limited by fatigue, subjectivity, and inconsistent judgement, making it difficult to assess defect boundaries or measure severity with confidence. DC Wheel Inspect was developed to bring greater objectivity, repeatability, and speed to this process by detecting and evaluating defects before components move further into assembly.

Let's talk

Combining zone-based vision analysis and AI inference for repeatable inspection at line speed

DC Wheel Inspect is an end-to-end machine vision system built around a fixed inspection station, where each steering wheel is positioned on a dedicated platform and captured from above by a vision system. The solution uses a hybrid inspection architecture that combines classical computer vision with AI-based inference to match the practical requirements of the application. Classical vision is used to segment the wheel into multiple inspection zones, each assigned a different level of structural criticality so that defects in more sensitive regions carry greater weight. An AI model then detects and localises subtle defect patterns that traditional thresholding methods cannot separate reliably from normal surface variation. Once a defect is identified, colour image processing is applied to measure intensity and generate a severity score, enabling objective quality assessment at production speed.