4. Camera placement and field of view

In high-speed occupant safety testing, camera placement is not a secondary production decision — it is a measurement decision. Every camera defines a specific measurement volume, and anything outside that volume is effectively unobserved. A test that produces beautiful 1,000 fps footage of the driver's head but misses the knee bolster interaction has failed to capture the data the test was designed to generate. This section covers how to reason about interior and exterior mounting, how to build rigid camera support structures on sleds and roll cages, and how to prevent occlusion during airbag deployment.

Defining the Measurement Volume Before Choosing a Camera

The starting point is always the occupant kinematics you need to quantify. For a frontal 56 km/h sled test with a Hybrid III 50th percentile male, the critical events typically include: airbag module cover opening (0–15 ms), bag membrane contact with the chest (25–45 ms), head rotation and possible face contact (50–90 ms), and knee bolster loading (30–70 ms). Each event occupies a different region of space, and the camera must see that region without obstruction for the entire duration.

A practical approach is to sketch the occupant's swept volume in the vehicle coordinate system. Using the standard SAE J670 convention — X forward, Y to the driver's right, Z upward, origin at the vehicle reference point — mark the head center of gravity travel, the chest deflection measurement location, and the knee target points. The camera's field of view (FOV) must then enclose this volume with margin. For a typical frontal sled, a head CG excursion of 350–450 mm forward and 150–250 mm downward is common, so the FOV should cover at least 600 mm × 500 mm in the sagittal plane.

Interior Mounting Strategies

Interior cameras fall into three functional categories: on-board occupant cameras, footwell/knee cameras, and airbag deployment cameras. Each has distinct constraints.

On-board occupant cameras are typically mounted on the B-pillar, roof rail, or a dedicated cross-car beam behind the front seats. The B-pillar position gives a near-lateral view of the head and chest, which is excellent for measuring head rotation and chest deflection. However, the B-pillar itself can occlude the shoulder belt path during pretensioner firing. A roof-rail mount looking down and forward captures head-to-airbag contact well but compresses the depth dimension, making out-of-plane motion harder to quantify. The cross-car beam position — roughly 800–1,000 mm behind the occupant — provides a clean three-quarter view that balances head, chest, and knee visibility.

Footwell and knee cameras are usually mounted low, near the sill or on the seat frame, looking forward and slightly upward. The challenge here is that the instrument panel lower edge and the steering column can block the view of the knee bolster. A useful technique is to mount the camera slightly outboard of the knee target and angle it inward, accepting some perspective distortion in exchange for an unobstructed line of sight.

Airbag deployment cameras require special consideration. The bag inflates in 20–40 ms, and the camera must capture the cover tear seam, the bag unfolding, and the initial membrane contact. A camera mounted on the steering column or the instrument panel above the module can be destroyed by the deploying bag. The standard solution is a sacrificial mount: a lightweight bracket that holds the camera in position but is designed to break away or be pushed aside by the bag without damaging the camera body or lens. The bracket should be positioned so the camera looks along the bag's deployment axis, not across it, to minimize the chance of the bag striking the lens directly.

Exterior Mounting Strategies

Exterior cameras serve two purposes: documenting the overall vehicle motion and providing orthogonal views that complement the interior cameras. For a sled test, the most valuable exterior views are the side view (Y-axis) and the top-down view (Z-axis). The side view captures the occupant's forward excursion and the airbag's deployment trajectory. The top-down view reveals lateral head motion and any asymmetry in bag inflation.

Exterior cameras are typically mounted on rigid stands or overhead trusses. The key requirement is that the camera and its mount must be isolated from the sled's reaction forces. If the camera stand is bolted to the same floor that the sled decelerates against, the stand will vibrate, and the footage will be unusable. Use separate foundation blocks or vibration-isolated mounts. A common specification is a mount natural frequency below 10 Hz, well below the sled pulse content, so the mount attenuates rather than amplifies vibration.

Roll Cage and Sled Fixture Design

On a roll cage or sled buck, camera mounts must be rigid enough to prevent relative motion between the camera and the occupant reference frame. A flexible mount introduces apparent motion that is indistinguishable from actual occupant motion. The rule of thumb is that mount deflection should be less than 1/10 of the smallest feature you need to resolve. If you need to resolve 5 mm of chest deflection, the mount should deflect less than 0.5 mm under the expected inertial load.

Practical mount design uses 25 mm × 25 mm steel or aluminum square tube, triangulated to the cage structure. Avoid cantilevered mounts; if a cantilever is unavoidable, keep the overhang under 150 mm and use a gusset. Camera mounting plates should be at least 6 mm thick and bolted with at least four M8 fasteners. For sled bucks, the camera mount should be attached to the buck structure, not to the sled carriage, so that the camera moves with the occupant rather than with the decelerating sled.

Avoiding Occlusion During Airbag Deployment

Occlusion is the most common cause of unusable high-speed footage in occupant safety testing. It occurs when an object — the airbag itself, the steering wheel, the seat, or a camera mount — blocks the line of sight between the camera and the measurement target. The airbag is the most dynamic occluder because it moves through the field of view during the critical measurement window.

To avoid occlusion, perform a line-of-sight analysis before the test. For each camera, trace the ray from the lens center to the target point at the time of interest. If the ray intersects any object that will be present at that time, the target is occluded. This analysis should be done at multiple time points: t = 0 ms (pre-deployment), t = 20 ms (mid-deployment), and t = 50 ms (post-deployment). A simple way to do this is to use a laser pointer mounted at the camera position and check whether the beam reaches the target at each stage of a static mock-up.

When occlusion is unavoidable, the options are: move the camera, change the camera angle, or use a mirror. Mirrors are particularly useful for seeing behind the airbag or around the steering wheel. A first-surface mirror mounted at 45° can redirect the line of sight without introducing the ghosting that occurs with second-surface mirrors. The mirror must be rigidly mounted and its position recorded so that the image can be geometrically corrected during analysis.

Practical Verification: The Static Mock-Up Check

Before committing to a camera layout, build a static mock-up of the occupant and the deployed airbag. Use a foam or cardboard representation of the bag at its maximum envelope, and place a laser pointer at each camera position. Walk through the following checks:

1. With the bag in the pre-deployment position, verify that each camera sees its target.
2. Replace the bag with the maximum-envelope mock-up and re-check. Any target that disappears is occluded during deployment.
3. Check the mirror paths if mirrors are used. Verify that the mirror surface is clean and that the reflected ray reaches the target.
4. Record the camera positions and angles in the vehicle coordinate system. These values become the input for the photogrammetry or DIC analysis that follows.

A useful rule for the mock-up is to add 20% margin to the airbag envelope. Real bags often inflate slightly beyond the nominal CAD envelope due to fabric stretch and seam geometry. If the mock-up shows a marginal line of sight, the real test will likely show occlusion.

Documentation and Repeatability

Camera placement must be documented with enough precision that the test can be repeated. Record the camera position in vehicle coordinates (X, Y, Z), the pan, tilt, and roll angles, the lens focal length, the sensor size, and the frame rate. For sled tests, also record the camera mount type and the torque on the mounting bolts. A simple table works well:

These values are illustrative and must be adapted to the specific buck geometry and occupant size. The key point is that the documentation must be complete enough that a different engineer can reproduce the same view on a different day.

Camera placement is ultimately a compromise between coverage, rigidity, and occlusion avoidance. The best layout is the one that captures the required measurement volume with the fewest cameras, each mounted rigidly enough that its own motion is negligible, and each positioned so that the deploying airbag does not block the critical line of sight. Verify the layout with a static mock-up, document it precisely, and the high-speed footage will yield data rather than just images.