During composites manufacturing, release coatings are applied on production tools to minimize tool-part friction and adhesive bonding. In addition to facilitating the removal of cured parts, applying release coatings reduces process-induced deformations (PIDs). The aerospace industry typically uses semi-permanent release coatings that undergo physical and chemical changes (i.e., aging) with each processing cycle. Fresh layers are frequently reapplied on top of aged coats to mitigate aging effects. Due to a lack of knowledge on the relationship between processing and aging, reapplications of release coating are often untimely and consequently lead to cost-deficient tool preparation schedules, or excessive PIDs. This paper presents a novel machine-learning (ML) framework to evaluate the condition of release coating using non-destructive and portable Fourier-transform infrared spectroscopy (FTIR) and contact angle goniometry (CAG). ML methods and other numerical tools are used to connect measurements gained from low-precision portable equipment to results obtained from high-precision laboratory instruments. The accuracy and portability of the technique demonstrates potential for scale-up and implementation into an automated industrial process. The research results contribute to understand the aging mechanisms of release coating, improve the efficiency of tool cleaning and preparation, and potentially mitigate PIDs in composites manufacturing.