Explained / Glossary
The words behind
the measurements.
Short, consistent definitions for the reliability, autonomy, evidence and worksite terms used across HumanoidUptime.
Showing all 32 terms
Reliability and measurement
14 terms- Availability / uptimeAlso: uptime
The share of a defined scheduled period in which a system was able to perform the required function. The time boundary, required function and rules for counting downtime must be stated.
Why it mattersOperating hours alone do not reveal how much planned work time was lost.Read the full guide →- Runtime / operating hoursAlso: operating time
Time accumulated while a robot is operating under the source's stated rules. It may be summed across several robots and does not necessarily include pauses or downtime.
Why it mattersA large runtime total shows exposure or activity, but cannot by itself produce an uptime percentage.Read the full guide →- Scheduled operating timeAlso: planned time
The time during which the robot was expected and available to perform the defined task.
Why it mattersIt supplies the denominator needed to interpret downtime and calculate operational availability.Read the full guide →- Downtime
Scheduled time in which the system could not perform the required function. A useful disclosure states whether planned maintenance, charging, waiting and external stoppages are included.
Why it mattersDifferent downtime rules can make two percentages look comparable when they are not.Read the full guide →- FailureAlso: failure event
An event in which the robot or system can no longer perform the required function under the stated conditions.
Why it mattersFailure counts are meaningless unless the required function and event boundary are defined consistently.- RecoveryAlso: restoration
The work and elapsed time required to return the system to the defined operational state after a stop or failure.
Why it mattersA system that stops infrequently can still disrupt production if each recovery takes a long time.- MTBFAlso: mean time between failures
Mean time between failures: the average operating time between one defined failure and the next in a repairable system.
Why it mattersThe figure depends on consistent failure rules, enough observed operating time and a clearly described system boundary.- MTTRAlso: mean time to repair, mean time to restore
An average time associated with restoring operation after a failure. The acronym is used for repair, restoration or recovery in different contexts, so the included steps must be defined.
Why it mattersDiagnosis, waiting for parts, physical repair, restart and verification can create very different recovery durations.- Human interventionAlso: intervention
A human action needed to keep, restore or redirect the robot's work, such as a reset, remote correction or physical assistance.
Why it mattersA workflow may produce useful output while still requiring frequent human help.- Observation windowAlso: measurement period
The dates, shifts or operating hours over which a result was observed and counted.
Why it mattersA result measured for one hour, one week or one year supports very different confidence about repeated operation.- Denominator
The total against which a count is interpreted: for example scheduled hours for an availability rate or all attempts for a task-success rate.
Why it mattersA count without its denominator cannot be converted into a meaningful rate.Read the full guide →- Task outputAlso: completed units
The number of reported items moved, parts handled, cycles completed or other units produced.
Why it mattersOutput shows useful work, but without attempts, rejects, retries and elapsed time it is neither a success rate nor a reliability rate.- Throughput
Task output per unit of time, such as tote moves per hour.
Why it mattersThroughput describes work rate under stated conditions, not how consistently that rate was sustained.- Fleet sizeAlso: unit count
The number of robots included in a reported result or observation.
Why it mattersA cumulative total across many robots cannot be interpreted as the experience of one robot unless the unit distribution is disclosed.
Autonomy and evidence
11 terms- Autonomous operationAlso: autonomy
Task execution by the robot without a person continuously issuing the individual control actions. The system may still be supervised and may request help.
Why it mattersThe word autonomous does not reveal how often people intervene, monitor, recover or remotely control the robot.- Supervision
Human monitoring of robot work, with the person available to approve, redirect or stop it when needed.
Why it mattersSupervised autonomous work can be valuable, but it has a different labor and scaling profile from unattended operation.- TeleoperationAlso: remote operation
A person controls some or all robot actions from a distance rather than the robot selecting those actions by itself.
Why it mattersA teleoperated demonstration proves physical capability, but not autonomous task performance.- Assisted cycle
A task cycle completed with human input beyond the workflow's stated normal operating mode.
Why it mattersCounting assisted and unassisted cycles together can hide how much human support the output required.- Operator corroborated
A deployment site, customer or operator publicly supports the specific claim.
Why it mattersIt adds evidence from the organization using the robot, but it is not the same as an independent controlled measurement.Read the full guide →- Manufacturer reported
The robot maker publicly reports the result and no reviewed operator or independent source confirms that specific figure.
Why it mattersThe claim may be useful evidence, but its provenance must remain visible.Read the full guide →- Independently measured
A researcher or test body outside the manufacturer and operating customer measured the stated result and published enough context to identify what was tested.
Why it mattersIndependent measurement can reduce source bias, but a lab result still does not automatically establish field reliability.Read the full guide →- Capability claim
A statement about what a product is designed or advertised to do without evidence that the result occurred in the deployment being reviewed.
Why it mattersA specification or demo claim should not be treated as a measured field result.Read the full guide →- PilotAlso: proof of concept, POC
A bounded trial used to test feasibility, integration or performance before broader operational adoption.
Why it mattersA pilot can provide real worksite evidence without proving long-term or scaled operation.- Field deploymentAlso: field use
Robot use in a real operating environment on a defined task, beyond a laboratory-only test.
Why it mattersField context is necessary for operational evidence, but the label alone says nothing about duration, scale or reliability.- BenchmarkAlso: controlled test
A test with specified procedures and conditions intended to measure one or more characteristics consistently.
Why it mattersA benchmark supports the characteristics it measures; it should not be generalized to untested jobs or environments.
Worksite vocabulary
7 terms- ToteAlso: tote box, bin
A reusable container used to move goods or components through a warehouse or factory.
Why it mattersA count such as 100,000 totes moved is task output, not operating time, unique-item count or a reliability rate.- Fixture
A device that holds or positions a workpiece so a manufacturing step can be performed consistently.
Why it mattersInserting a part into a fixture is a specific task outcome; it does not describe the whole production process.- Sequencing trolleyAlso: sequencing cart
A mobile rack or cart loaded with parts in the order required by a later production step.
Why it mattersLoading or unloading one is a logistics task whose success conditions depend on the plant workflow.- Body shop
The part of an automotive plant where metal body components are joined and assembled before painting and final assembly.
Why it mattersThis is an industrial production environment, not a retail workshop or vehicle repair shop.- Intralogistics
The movement, storage and coordination of materials within a factory, warehouse or site.
Why it mattersMany current humanoid deployments move containers or components rather than manufacture the product itself.- Pick and place
A task in which a robot grasps an item at one location and places it at another defined location.
Why it mattersSuccess depends on the item, presentation, destination, speed and error rules used in the test.- Battery swapAlso: autonomous battery exchange
Replacement of a depleted battery with a charged battery, performed manually or by the robot itself.
Why it mattersAutomating the swap can reduce charging interruption, but does not by itself prove autonomous task execution or continuous availability.
No term matches that search.
Need the method?
Definitions are
only the beginning.
See how HumanoidUptime applies these terms when reviewing public evidence.
Read the methodology