Inside the Data Bottleneck Slowing Visual and Physical AI

A 2026 survey of more than 700 professionals documents how teams build visual and physical AI, according to a white paper published by IEEE Spectrum and Wiley and sponsored by Voxel51.

The report states that 78% of teams already see measurable value from visual and physical AI, while 74% still consider the field underinvested relative to its opportunity. It also reports that teams which ship successfully invest nearly 3x more time in data work than teams that struggle, and that 92% of practitioners believe the field is heading somewhere specific next.

The findings indicate that data problems cause the majority of model failures and that curating data matters more than chasing larger architectures. Annotation remains costly and wasteful, the report says, because teams often label everything and then discard much of it before production. Data work, not data collection, separates teams that ship from teams that stall, according to the findings.

The report covers systems driven by video, LiDAR point clouds, sensor streams and other high-dimensional data that perceive, reason and act in physical space. The white paper is available through a registration and login process on the hub.