Physical AI Benchmark Database
Compare published Physical AI benchmark results while keeping robot, environment, trial count, protocol and adaptation conditions attached to every score.
Explore datasetTechniaHQ Research Library
Explore collected robotics records, compare reported results and follow the original evidence.
This library contains records from different publication dates. It is not a live industry feed.
Page updated November 11, 2026 · Individual source dates appear within each dataset
Browse the archive
19
Linked data views
105
Structured records in /data
Find a dataset
Search by topic, robot type or dataset name. Each collection links to its own records, filters and sources.
01 / Physical AI evidence
Benchmarks, real-robot tests, training data, inference latency, deployments and documented failures.
Compare published Physical AI benchmark results while keeping robot, environment, trial count, protocol and adaptation conditions attached to every score.
Explore datasetSee what Physical AI models actually demonstrated on real robots, including task, control mode, trials, failures, interventions and evidence limits.
Explore datasetCompare six reported robot-learning data budgets by training stage, unit, included subsets and access limits. Counts are reported by the authors, not independently recounted.
Explore datasetCompare published robot-model inference latency with GPU, precision, model size, batch size and test conditions kept visible beside every result.
Explore datasetTrack documented Physical AI and embodied robotics deployments while separating pilots, agreements, operating sites, model disclosure and intervention evidence.
Explore datasetReview documented Physical AI task failures and public failure taxonomies without guessing causes or turning isolated events into universal reliability claims.
Explore dataset02 / Robotics labor market
Recorded hiring activity, requested skills, published salaries and research-to-industry links.
Compare retained Physical AI hiring activity by company using active listings and role mix. The index measures public hiring activity, not workforce size.
Explore datasetMeasure which skills are explicitly requested in retained Physical AI job descriptions, including programming languages, ML methods, simulation, perception and robot control.
Explore datasetCompare only publicly disclosed compensation from retained Physical AI and robotics job listings, preserving original currency and annual, monthly or hourly periods.
Explore datasetTwo historical Agility Robotics appointments, separating roles taken in May 2023 from the leadership changes announced on 15 May 2024.
Explore datasetFour sourced connections involving a robotics spinout, a researcher's education and published model collaborations. These are different relationships, not a university ranking.
Explore dataset03 / Humanoid intelligence
Robot specifications, comparisons, deployments, shipments and reported failures.
Connected evidence for humanoid models, deployments, failures, production, shipments and funding.
Explore datasetSearch retained humanoid robot specifications, manufacturers, prices and commercial status.
Explore datasetCompare source-backed humanoid specifications without filling unknown fields with estimates.
Explore datasetSource-linked charts on prices, hardware, availability and reliability evidence.
Explore datasetTrack Chinese humanoid models, manufacturers, published prices and commercial status.
Explore datasetReview production milestones, shipments and deliveries without merging these categories.
Explore datasetDocument operating sites, customers, tasks, deployment status and disclosed unit counts.
Explore datasetReview recorded failures, operator interventions and denominators when reported.
Explore datasetReading the records
Each dataset reports what its retained sources establish. Dates, scope and test conditions affect comparisons.
Open the original reference. A listing or specification recorded months ago may have changed.
Robot tests require task descriptions and sample sizes. Hiring counts refer to retained listings, not a company’s workforce.
Unknown measurements remain undisclosed. Production, shipments and deployments are counted separately.
Read the Physical AI guide before comparing models or return to the robotics collections.