Why "in-the-loop" is the most misused phrase in simulation.
A USB cable is not a loop. A motion seat reacting to telemetry is not a loop. A projector moving beside a cockpit is not a loop. Those are connected accessories. A true loop is the closed biological-mechanical feedback circuit between physics, sensation, interpretation, control, and new vehicle state.
"The loop is the closed, continuously updating cause-and-effect chain between vehicle physics, sensory delivery, human interpretation, human control input, and resulting vehicle state change."
This definition is precise by necessity. Every word carries structural weight. The loop is not a marketing label for motion hardware. It is not shorthand for any simulator that moves. It is a technical description of a specific kind of closed feedback relationship between a physical system and a biological one.
The chain has five components. Each must be causatively connected to the next. If any link is absent, delayed beyond the neurological threshold, or substituted with an approximation, the loop is broken at that point regardless of how convincing the experience appears.
"The loop is not the connection between simulator components. The loop is the uninterrupted connection between simulated vehicle physics and human neurological response."
Component connectivity is an engineering condition. It can be achieved with wiring and software. Neurological integration is a biological condition. It requires causative accuracy, temporal coherence, and sufficient signal strength. These are not the same requirement, and meeting the first does not imply meeting the second.
The loop operates as a continuous cycle. Each node feeds the next. Interrupting or degrading any connection breaks the full chain.
Physics creates sensation. Sensation creates interpretation. Interpretation creates control. Control changes physics.
Component connectivity and neurological closure are different conditions. A system can have all hardware connected and still be structurally open-loop in the sense that matters for driver training.
| Connected System | Closed Loop System |
|---|---|
| Hardware attached to software | Human nervous system embedded inside the physics event |
| Motion may react after the event has passed | Motion is causative and time-coherent with the event |
| Visual-first interpretation: driver reads the screen | Vestibular-first interpretation: driver feels the vehicle |
| Driver reacts to a presentation of vehicle behavior | Driver responds naturally to vehicle state as it evolves |
| Open-loop entertainment behavior | Closed neurophysical training behavior |
The distinction is not about how impressive the system looks or how many channels of motion it produces. It is about whether the human participant is a passive observer of simulation output or an active element inside the simulation's causal chain.
Genuine closed-loop participation requires three conditions to be met simultaneously. All three must hold. Meeting two is not sufficient.
The sensory cue must come from the real simulated physics event, not from a canned effect, visual trigger, or decorative motion profile. If the motion is scripted rather than derived from live rigid-body state, the causal chain is broken at the source.
The visual, vestibular, haptic, audio, and control channels must arrive in the correct order and timing relationship. If yaw is late, visuals lead motion, or force feedback arrives out of sequence, the loop is broken regardless of individual channel accuracy.
The cue must be truthful and strong enough to alter natural control behavior. If the driver is still primarily driving by sight because the body cue is insufficient, the human is not truly in the loop. Signal strength is not optional.
The three fundamental criteria define what the loop requires at a conceptual level. The six supporting requirements in the In-the-Loop Standard define how those criteria are structurally satisfied. The criteria are the standard; the requirements are the structural test. Both must be consulted together for a complete in-the-loop determination.
| Fundamental Criterion | Supporting Requirements (In-the-Loop Standard) |
|---|---|
| A — Causative AccuracyMotion cue originates from live physics state | Req. 1: Physics-Driven MotionReq. 2: Center-of-Mass Reference |
| B — Temporal CoherenceSensory channels arrive in correct order and timing | Req. 3: Independent Degrees of FreedomReq. 4: Real-Time System Coherence |
| C — Human Response RelevanceCue is sufficient to alter natural control behavior | Req. 5: Vestibular RelevanceReq. 6: Training Validity |
Meeting a criterion requires satisfying all of its supporting requirements. Partial satisfaction does not constitute in-the-loop status.
The canonical definitions for all three criteria are maintained in Canonical Definitions.
Each system architecture has a characteristic failure mode. Understanding where the break occurs is more useful than a categorical pass/fail label.
The vestibular channel is absent. Visual information replaces physical vehicle state cues entirely. The driver builds spatial awareness from screen content rather than from body sensation. Control responses develop as visual reactions, not as proprioceptive corrections. The human is positioned outside the physics event.
The body is moved relative to the cockpit rather than the cockpit and body moving together inside the vehicle's center-of-mass behavior. This produces a sensation of the seat moving under the driver rather than the vehicle rotating around a shared physical reference. The inertial signal does not match the vehicle's actual rigid-body event.
Cueing is disturbance-based and actuator-stacked rather than true center-of-mass vehicle motion. The system reconstructs vehicle behavior from post-processed effects applied at the contact points rather than generating motion from the rigid-body physics state directly. The resulting cue approximates some aspects of vehicle behavior but is not causatively derived from the event it represents.
Every actuator participates in every cue, creating a dependent architecture. Because all six actuators contribute simultaneously to yaw, pitch, roll, surge, sway, and heave, delivering accurate independent timing for each axis is structurally constrained. The system cannot preserve the correct temporal relationship between independent vehicle motion axes when multiple events occur simultaneously, which is the normal operating condition at the limit of vehicle performance.
Preserves the loop by allowing the driver, cockpit, control interface, and visual environment to respond as one causative vehicle event. Each degree of freedom is controlled independently, preserving the correct timing relationship between axes. The center-of-mass origin of motion ensures the inertial reference matches the vehicle's actual rigid-body behavior. The driver is placed inside the physics event rather than in front of a representation of it.
The Simulation Fidelity Rating framework is structured around a single primary question:
"How complete is the human closed-loop participation?"
The question is not how many actuators a system has. The question is not whether it moves. The question is whether the participant's sensory nervous system receives the same causative information at the same relative moment it would in the real vehicle, so that control correction happens inside the same evolving event.
SFR scoring reflects the degree to which a system meets all three requirements: causative accuracy, temporal coherence, and human response relevance. A system that scores high on one requirement but fails another cannot claim full closed-loop status regardless of the impression it produces.
A simulator is in-the-loop only when the participant's sensory nervous system is receiving a causatively accurate and time-coherent translation of the simulated vehicle's actual rigid-body state, such that their control corrections occur as part of the same evolving physical event.
For any simulation system under evaluation, answer each question. A single "No" indicates the system is not fully in-the-loop for that condition.
Check each condition that is met. Unchecked items identify the specific points where the loop is broken.