Thermal Interface Materials Inspection
Ensure TIM correctly applied with deep learning solutions
Batteries can generate a lot of heat, which must be removed to prevent battery damage or premature performance decline. Thermal interface materials (TIM) are used to conduct heat away from the battery. Many TIMs simultaneously serve the equally important function of electrical insulation.
TIMs must be applied precisely, with close contact between substrates. A wide range of defects, including air bubbles, poor adhesion, and inclusions, can reduce both thermal conduction and electrical insulation. Visual inspection must identify a wide range of possible flaws in installation and application, often involving materials with poor color contrast. Once the battery assembly moves to the next step, the TIM is permanently concealed and unavailable for further inspection. Errors at this stage can lead to hard-to-diagnose problems down the line.
While rule-based machine vision can accurately detect anticipated problems with beads, gaps, installation width, and other common features, Cognex Deep Learning learns to detect a significantly wider array of installation problems with every type of TIM. If a battery should later fail, its failure mode can be tied back to a specific stored image of the TIM, and the deep learning training model can be further refined to detect these new errors.