When autonomy needs a human,
is the human ready?
HARP is an open operational architecture prototype for measuring the strength and recency of competency evidence among people supporting safety-critical autonomous systems.
Autonomy changes where humans enter the system.
As autonomous systems handle more routine operation, human involvement can become less frequent but more consequential. Traditional training records answer whether someone completed training or holds a qualification. HARP explores a different question: how strong is the current evidence that required competencies can still be demonstrated?
From training completion to operational evidence.
Remote Assistance Simulator
Work scripted edge cases, gather context, choose bounded actions, receive deterministic after-action feedback, and create new competency evidence.
Launch simulator →Ambiguous Construction Flagger
Qualified headcount does not guarantee competency coverage.
Readiness beyond headcount.
Roll individual evidence into shift-level coverage signals, competency heatmaps, requalification queues, and targeted training priorities.
Open readiness dashboard →Qualification ≠ readiness.
An operator can remain formally qualified while the evidence supporting a rare, safety-critical competency becomes stale. HARP keeps those concepts separate instead of pretending a course-completion date proves present-day proficiency.
“The less frequently humans are required to intervene, the more important it becomes to know they are ready when intervention is necessary.”EXPERIENCE HARP →