# Hexagon OPTIV S Signals a Phased Move Toward AI-Assisted Optical Measurement https://cncoutlook.com/news/hexagon-optiv-s-optical-cmm-2026/ By CNC Outlook Editorial · 2026-09-07 OPTIV S combines optical and selected tactile measurement methods while Hexagon introduces AI-assisted functions over time; available and planned capabilities must be distinguished. An AI feature promised for a future software release is not an instrument capability available at installation. That distinction matters for Hexagon's [OPTIV S]() optical CMM, introduced with a phased roadmap for software assistance alongside its multisensor inspection architecture. Hexagon also describes a roadmap of AI-assisted functions being introduced in phases during 2026 and 2027. That timeline matters for a purchasing decision. A capability demonstrated in a future-release preview should not be treated as something already available in every installed system. Availability is especially important with phased software releases. A manufacturer purchasing the instrument in late 2026 should obtain a list of functions included at delivery and the roadmap for planned AI capabilities. Licensing, support periods and controller requirements may influence whether future features can be adopted without significant cost or downtime. Clear release documentation matters more than a broad statement that the product is AI-enabled. This is particularly true where a new function could alter an established inspection procedure. The new platform combines optical measurement with sensor options that can extend its use beyond image-based inspection. Different sensors collect different information, and the measurement program must choose the approach that can establish a characteristic reliably. A small hole, a soft surface and a reflective machined plane do not necessarily suit the same method. ## Optical and tactile sensors share one datum system Optical measurement depends on illumination, contrast and part presentation. A camera can detect a well-defined edge readily on one material yet struggle with a dark, textured or reflective surface. Multi-sensor configurations offer alternatives, but require calibration and appropriate coordinate registration between the sensing methods. A mixed-sensor measurement plan must be consistent about reference geometry. If a tactile probe establishes a datum plane and an optical system measures a nearby profile, those sensors need to share a calibrated coordinate relationship. Differences in temperature, lighting, or the part's clamping state can affect the final calculation. When machine software automatically selects measurement strategies, a quality team should still be able to review the sensing method and the assumptions behind each reported characteristic. The delivery configuration—not the roadmap—must determine the qualified inspection procedure. ## Where promised AI meets inspection approval For a CNC supplier, the decision often concerns inspection throughput as much as an absolute accuracy claim. If an optical machine can automatically measure repeated small components, it may relieve operators from manual inspection. But programming, loading, cleaning and report generation can still dominate total cycle time. Users should request the exact configuration and software release supported at delivery. A representative part test should show the accepted feature measurements, repeatability, illumination conditions and how the machine handles poor edge detection or sensor changes. Where AI assists feature recognition, there must be a clear way to review and override an incorrect suggestion. For an OPTIV S buyer, the strongest specification is a list of supported sensors, validated features and installed software functions at the chosen delivery date. Future AI assistance may prove helpful, but it should not be treated as an unverified promise of faster or more accurate measurements today. ## Sources & References - [Hexagon OPTIV S Brings Multisensor Options to Optical CMM Inspection]() — Metrology and Quality News (2026-09-07T12:00:00Z) © CNC Outlook. See https://cncoutlook.com/terms/ for reuse terms.