3. CRS classification and installation modes
Child restraint system (CRS) fit assessment begins with two questions that must be answered before any measurement is taken: what class of restraint is this, and how is it attached to the vehicle? The answers determine which reference planes, adjustment ranges, and belt-routing checks apply. A seat that is correctly classified but installed through the wrong path will fail fit criteria even when its nominal size range appears adequate. This section establishes the classification taxonomy and the installation-mode distinctions that drive every downstream fit decision.
Classification by child size and restraint strategy
CRS products are grouped by the physical mechanism they use to manage the child's motion in a crash, not by marketing labels. Four families dominate the market:
- Infant carriers (Group 0/0+) — rear-facing only, typically rated from birth to 13 kg (approximately 0–15 months). The shell is a semi-enclosed bucket with an integral or detachable base. The child is held by a three- or five-point harness, and the carrier is oriented so the child's back faces the direction of travel.
- Convertible seats — a single shell that can be installed rear-facing and later re-oriented forward-facing, usually covering roughly 0–18 kg rear-facing and 9–25 kg forward-facing depending on the model. The recline angle and harness routing change between modes, so the same product presents two distinct fit problems.
- Combination and booster seats — forward-facing harnessed seats that convert to belt-positioning boosters, or dedicated boosters rated from about 15 kg upward. Boosters do not restrain the child directly; they reposition the child so the vehicle's three-point belt crosses the correct skeletal landmarks.
- Special-needs and extended-harness restraints — niche products with higher weight limits or postural support features, assessed with the same classification logic but additional harness and support criteria.
Classification matters for fit because each family defines a different occupant reference geometry. An infant carrier is assessed with the child reclined and the head contained within the shell; a booster is assessed with the child upright and the belt routed across the shoulder and pelvis. Applying infant-carrier recline criteria to a booster, or booster belt-fit criteria to a rear-facing shell, produces meaningless results.
Installation modes: ISOFIX/LATCH versus seat belt
Every CRS attaches to the vehicle through one of two primary load paths, and some products support both.
ISOFIX / LATCH (Lower Anchors and Tethers for Children) uses two rigid or semi-rigid connectors that engage the vehicle's lower anchor bars, plus a top tether or support leg to control rotation. The connection is mechanically keyed: the connector either clicks into the anchor or it does not, which removes much of the belt-path ambiguity. However, ISOFIX is not universal — anchor spacing, anchor depth, and the presence of a top-tether anchorage vary by seating position and vehicle generation. A CRS may be ISOFIX-approved but physically blocked by a deep seat cushion or a protruding buckle stalk.
Seat belt routing uses the vehicle's three-point belt threaded through a specific path on the CRS — typically a rear-facing belt path behind the shell, a forward-facing path through the backrest, or a lock-off clamp. Belt routing is more sensitive to vehicle geometry: belt buckle length, stalk stiffness, retractor behavior, and the angle at which the belt leaves the CRS all affect whether the installation is tight and whether the belt path is fully engaged.
How installation mode changes the fit problem
The installation mode determines which degrees of freedom remain adjustable after the seat is fixed. With ISOFIX, fore-aft position is largely set by the anchor location, and the assessor's remaining levers are recline angle, headrest height, and harness fit. With belt routing, the assessor can often shift the CRS fore-aft within the belt's reach, which changes legroom, recline, and the child's distance from the front seatback. These differences propagate into the fit criteria:
| Fit criterion | ISOFIX/LATCH installation | Seat belt installation |
|---|---|---|
| Installation tightness | Connector engagement and top-tether/support-leg contact | Belt tension, lock-off engagement, buckle clear of shell |
| Recline adjustment | Base or foot adjustment within anchor geometry | Belt path and recline foot jointly set the angle |
| Fore-aft position | Fixed by anchor bar location | Variable within belt reach; affects legroom and front-seat clearance |
| Common failure | Anchor blocked, connector not fully latched, tether not routed | Belt twisted, buckle stalk contacting shell, insufficient tension |
In practice, the assessor should record the installation mode explicitly on the fit record, because a "pass" obtained in one mode does not transfer to the other. A convertible seat that fits a child well when belt-installed at a given recline may sit at a different angle when ISOFIX-installed, changing head containment and harness geometry.
Worked example: classifying and checking a convertible seat
Consider a convertible seat rated 0–18 kg rear-facing and 9–25 kg forward-facing, installed in the rear outboard position of a compact hatchback. The child is 14 months, 9.8 kg, 76 cm. The assessor first classifies the intended mode: rear-facing, because the child is below the forward-facing minimum and within the rear-facing weight limit. The installation mode is ISOFIX with top tether, and the vehicle's outboard anchors are present and unobstructed.
The assessor then checks the mode-specific criteria in sequence: connector latches fully engaged on both anchors (audible click, indicator green); top tether routed to the designated anchorage and tensioned; recline indicator showing the rear-facing zone appropriate for a child under 12 months (the seat's angle gauge reads within the marked band); harness height at or just below the child's shoulder; harness snug with no more than a flat-finger gap at the collarbone; head contained within the shell with the top of the head below the shell rim. Each item is recorded as pass, fail, or not applicable, with the installation mode noted. If the same seat were belt-installed, the assessor would additionally verify that the belt is routed through the rear-facing path, the lock-off is closed, and the buckle does not bear on the shell — checks that do not exist in the ISOFIX path.
Implications for the fit record
Because classification and installation mode jointly define the applicable criteria set, both must be captured before any dimensional or harness check. A fit record that lists only "convertible seat, rear-facing" without the attachment mode is incomplete: it cannot be reproduced, and it cannot be audited against the correct criteria. The practical rule is to treat class + mode as the primary key of the fit assessment, and to re-run the full criteria set whenever either changes — for example, when a child outgrows the rear-facing limit and the seat is re-oriented, or when the family moves the seat to a vehicle position without lower anchors.
The diagram above summarizes the decision structure: classification and installation mode are parallel inputs that converge on a mode-specific criteria set, with a shared child-fit layer applied in both cases. The key operational takeaway is that neither classification nor installation mode can be assumed from the product name or the hardware visible on the seat. Both must be read from the approval label and the vehicle's anchor configuration, recorded explicitly, and used to select the criteria that the fit assessment will actually apply.