Required and optional data inputs for SFR evaluations.
This document defines the data inputs required to conduct an SFR evaluation. It separates required inputs — without which classification cannot be made — from optional inputs that enhance analysis but are not necessary for a determination. Classification must remain achievable using required inputs alone.
SFR evaluation inputs are divided into two categories: required and optional. This distinction is structural, not aspirational.
A required input is one without which a criterion assessment cannot be completed. If a required input is unavailable, the relevant criterion cannot be assessed and the classification for that criterion defaults to Insufficient Data, not Pass. Optional inputs may enhance the analysis and support borderline determinations, but their absence does not automatically prevent classification.
This document is used in conjunction with the Reference Test Methodology (which defines what events are tested) and the Evaluation Process (which defines how inputs are assessed against criteria).
An evaluation conducted without all required inputs is incomplete. An incomplete evaluation produces Insufficient Data, not a classification.
Measured output of the motion system at the cockpit reference point. Must include position, velocity, and acceleration data on all actuated axes during each reference event.
The vehicle physics model output at the moment of each reference event. Must include vehicle position, velocity, acceleration, yaw rate, pitch rate, and roll rate as computed by the physics engine, time-stamped at the physics computation cycle.
Command and response data for each motion actuator during reference events. Must include the command signal sent to each actuator and the measured response, enabling comparison between commanded and delivered motion.
Timing data showing the relationship between the physics event timestamp, the motion system response timestamp, and the visual system frame timestamp. Must be captured from a common timing reference.
Response data from steering, throttle, and brake systems during reference events. Captures the participant's control corrections during each reference event, timestamped relative to the physics event that prompted the correction.
Captures gaze direction and fixation patterns during reference events. Most useful for Criterion C (Human Response Relevance): an evaluator who observes that the participant's gaze is fixed on a visual reference point at the moment of a limit-state event has additional evidence that the participant is not responding to physical sensation.
Physiological response data during reference events, including heart rate variability, galvanic skin response, and muscle activity patterns. Provides additional evidence of physiological stress response to physics events.
Neural response data correlated with reference event timing. Captures cortical processing patterns associated with vestibular versus visual stimulus processing. Requires lab-validated instrumented testing to apply reliably.
Precise measurement of the interval between physics event onset and participant control response initiation. Captures the reaction latency pattern across events and can distinguish between anticipatory visual responses and reactive physical sensation responses.
The following rules govern how input availability affects classification outcomes.
| Condition | Effect on Classification |
|---|---|
| All required inputs present and valid | Classification may proceed. Each criterion is assessed against its relevant inputs. Outcome may be Pass, Fail, or Insufficient Data per criterion. |
| One or more required inputs missing | Criteria that depend on the missing input default to Insufficient Data. Classification of the overall system is Insufficient Data if any required criterion cannot be assessed. Classification proceeds on criteria for which required inputs are available. |
| Required inputs present but below quality threshold | Treated as missing for classification purposes. Low-quality data is not substituted with optional inputs or lower-tier evidence to fill a required input gap. |
| Optional inputs absent | No effect on classification availability. The evaluation proceeds on required inputs. A note may be made that optional inputs were not available, but this does not change the classification outcome. |
| Optional inputs present and contradicting required inputs | Required inputs (Tier 1 evidence) take precedence. Optional inputs may prompt a notation of discrepancy in the evidence summary but cannot override a classification established from required inputs. |
Optional inputs cannot replace required inputs. Absence of optional inputs does not trigger Insufficient Data.
Required inputs must meet the following minimum quality standards to be considered valid for classification purposes. Data that does not meet these standards is treated as unavailable.
Specific minimum sample rates and calibration intervals will be defined in a supplementary technical specification document when the framework advances to v1.0.
The quality and completeness of evaluation inputs directly determines the quality of the classification outcome. A system that restricts evaluator access to required inputs cannot be fully classified. A system that is fully classified on required inputs alone has produced a valid, complete determination.
Optional inputs exist to strengthen and clarify borderline determinations, not to replace the required inputs that form the foundation of the classification framework. The five required inputs represent the minimum data set for a repeatable, credible SFR evaluation.