Observation
Preserve sensor coverage, calibration and disagreement.
System anatomy · 07
Perception, grounding and uncertainty
Connect observations, object state, instructions and action permission to a recoverable physical state estimate.
Evidence statusEngineering evidence framework · task and site validation requiredLast reviewed: 14 August 2026
07
FUURAA thesis
A language-plausible action is not a physically grounded or authorised action.
Engineering map
Turn the title into engineering objects that can be observed, measured and reviewed.
Each layer states the boundary to establish and the evidence needed for the next decision.
State
Represent object, tool, people and robot uncertainty.
Authority
Bind uncertain states to ask, wait, stop or act.
Verification questions
Write the questions first, then decide whether a demo, test, pilot or operating record can answer them.
Each question needs an object, conditions, denominator, threshold and accountable decision owner.
- 01
What cannot the robot observe?
- 02
How is uncertainty represented?
- 03
Which actions require confirmation?
- 04
Can later review reconstruct the decision?
Evidence to preserve
Enable the next reader to reconstruct conditions, results, failures and the decision.
A conclusion alone loses reviewability; raw records, configuration and exclusions matter too.
- 01
Evidence package 1
Sensor and state-coverage map
- 02
Evidence package 2
Uncertainty-trigger policy
- 03
Evidence package 3
Decision trace with physical outcomes
Scope boundary
State what this evidence still cannot be generalised to.
- 01
Model or perception accuracy on a dataset does not establish safe action in an open physical environment.
- 02
This is an engineering reading and decision framework; it does not replace application-specific risk assessment, conformity work, procurement acceptance or professional advice.
Sources and evidence status
Read standards scope, measurement evidence and application conclusions separately.
Source dates and review status remain visible; external sources open in a new tab.
01
NIST · Updated 2024-08-16 · verified 2026-08-14
↗02Standard test methods for response robots
Develops repeatable measures for mobility, manipulation, sensing, endurance, communications, operator proficiency, logistics and safety.
MIIT HEIS Committee / Beijing E-Town · Released 2026-02-28 · reported 2026-03-02
↗China's Humanoid Robotics and Embodied Intelligence Standards System (2026)
Organises work across common foundations, embodied models and computing, limbs and components, complete systems, applications, and safety.