Robot-Training Labor Market: Advertised Pay and Open Listings, September 21-22, 2026

Two dated observations of advertised hourly pay and open-posting counts in a selected set of public robot-training and physical-AI employer feeds. Compiled by RobotTraining.jobs from its public pay.json and roles.json aggregate endpoints on September 21 and 22, 2026.

The latest observation contains 687 open postings across 56 employers and seven work categories; 465 postings are attributed to TSMG. This is a selected sample, not a representative wage survey or a count of workers. The two days do not establish a hiring trend.

The archive includes four CSV tables (pay by work type, openings by work type, openings by employer, and totals), both raw JSON observations, an export and validation script, and a README with column definitions, provenance and limitations. Pay is emitted only with at least eight distinct rate samples from three employers. Missing or suppressed pay is left blank. Counts reconcile across both breakdowns and the totals.

Source collection uses rules and, where configured, language-model classification; category and pay-extraction errors are possible. Documentation and code were prepared with AI assistance, while all numerical observations derive from the cited public feeds. No applicant data or job-description text is included. The archive is a dated release, not an automatically synchronized live mirror.

Current aggregates and their source timestamps: tracked open listings and advertised hourly pay.