2
Manufacturer publications
VALIDATION CLAIMS AND DATA CATEGORIES
Inspect AgiBot walking-validation statements and five interaction-data categories without treating them as incident counts or fleet reliability rates.
2
Manufacturer publications
5
Interaction categories
No category frequencies published in the cited announcement
1
Paired count / exposure claim
Up to 1,300 walking hours without anomalies, according to AgiBot
| Robot | Failure / category | Reported count | Stated exposure | Intervention data | Evidence type | Reading limit | Source |
|---|---|---|---|---|---|---|---|
| AgiBot A2 Ultra AgiBot | Fall | Not publicly disclosed | No paired denominator established | Not disclosed | Manufacturer claim Manufacturer validation claim | The source lists a 5,100-hour walking test, then up to 1,300 continuous hours without anomalies and an equivalent typical-use zero-fall period. It does not clearly pair zero falls with 5,100 hours; no rate is calculated. | Open source |
| AgiBot A2 Ultra AgiBot | Operational anomaly | 0 | Up to 1,300 continuous walking hours | Not disclosed | Manufacturer claim Manufacturer validation claim | The current official product page states up to 1,300 hours of continuous walking without anomalies. | Open source |
| AgiBot World dataset AgiBot | Missed grasp | Not publicly disclosed | No paired denominator established | Not disclosed | Published data category Exploratory teleoperation category | The announcement names this interaction outcome. It supplies no event count or exposure denominator; this is not a field-incident record. | Open source |
| AgiBot World dataset AgiBot | Collision | Not publicly disclosed | No paired denominator established | Not disclosed | Published data category Exploratory teleoperation category | The announcement names this interaction outcome. It supplies no event count or exposure denominator; this is not a field-incident record. | Open source |
| AgiBot World dataset AgiBot | Object drop | Not publicly disclosed | No paired denominator established | Not disclosed | Published data category Exploratory teleoperation category | The announcement names this interaction outcome. It supplies no event count or exposure denominator; this is not a field-incident record. | Open source |
| AgiBot World dataset AgiBot | Unstable contact | Not publicly disclosed | No paired denominator established | Not disclosed | Published data category Exploratory teleoperation category | The announcement names this interaction outcome. It supplies no event count or exposure denominator; this is not a field-incident record. | Open source |
| AgiBot World dataset AgiBot | Liquid splash | Not publicly disclosed | No paired denominator established | Not disclosed | Published data category Exploratory teleoperation category | The announcement names this interaction outcome. It supplies no event count or exposure denominator; this is not a field-incident record. | Open source |
The walking statement is a manufacturer claim. The interaction dataset was collected through exploratory teleoperation, including intentionally guided contacts. Neither source is a fleet incident log, and no independent validation or video inspection is claimed here.
METHODOLOGY
Two manufacturer publications support this limited collection. Five rows name categories from exploratory teleoperation data, not counted field incidents. The walking page juxtaposes 5,100 hours, up to 1,300 continuous hours and a zero-fall equivalence; their denominators must not be merged. Source review: 29 September 2026.