Automate facial mocap setup — generate 52 ARKit shapekeys in Blender

Facial mocap is increasingly available via iPhone TrueDepth/ARKit, but studios and professional creators still spend hours converting raw blendshape streams into usable Blender rigs: mapping ARKit's 52 blendshapes to clean, topology-aware shapekeys, creating corrective targets, and producing reliable exports for UE5/Metahuman or offline animation remains manual, inconsistent, and error-prone for many of the ~120,000 studios and creators in the $3.6B capture/rigging tools market. The pain is concentrated at small to mid-sized teams that have adopted Blender pipelines but lack dedicated rigging specialists, and the inefficiency shows up as days of setup per character and fragile live pipelines during production. You could build a Blender-first automation product — an add-on plus CLI pipeline — that ingests ARKit/TrueDepth captures, generates the canonical 52 ARKit shapekeys, auto-creates corrective shapes and normalized neutral poses, validates topology/mirroring, and exports studio-ready FBX/GLTF/UE5 formats. An MVP would target reducing setup time from multiple days to minutes and aim for high pass rates on common topology classes, while surfacing cases that need manual intervention; technical challenges include handling wide topology variance, naming/standard compatibility, and producing high-quality correctives without per-face sculpting. This market looks attractive now because phone-based capture ubiquity, rising Blender adoption, and demand for real-time pipelines put a premium on repeatable, Blender-native tooling (market score 92/100, revenue potential 88/100). Competition is medium, so a clear differentiation — a one-click ARKit→Blender pipeline, tight UE5/Metahuman export, batch processing for studios, and transparent quality metrics — could win adoption, but success will require investing in a broad validation dataset, solid UX for failure cases, and partnerships with capture-app vendors to reduce friction.