THE TARGET COMMANDER LOOP
One evolving mission state.
STARHELM is designed so observation becomes understanding, understanding becomes choice, choice becomes coordinated action, and outcome becomes the next state.
- 01
Observation
Source, time, custody, and sensor context enter the mission thread together.
- 02
Estimated state
The operator sees the current estimate, uncertainty, and alternative interpretations.
- 03
Threat and intent
Mission context connects observed behavior to objectives, constraints, and likely response.
- 04
Courses of action
AI-assisted options branch against adversary behavior, mission objectives, and predicted consequence.
- 05
Human authority
The commander selects the course and commits authority at the consequential decision point.
- 06
Task
The mission application expresses bounded work for the customer-selected system.
- 07
Acknowledgement
Acceptance, refusal, and execution state return to the same operator thread.
- 08
Outcome
Observed results remain connected to the decision and task that produced them.
- 09
Assessment and replan
The new state becomes the basis for AAR, learned behavior, and the next command decision.
EVALUATION UNIT
Score the complete thread.
One representative mission. One operator-controlled decision loop. Evaluator-held truth at every consequential transition.
Discuss an evaluation →