Rear-Facing Infant Seat Testing and Research


Transport Canada does testing and research on rear-facing infant seats. We use the results to suggest changes to seat safety regulations. Since testing and research are always going on, we will update this Web page with any new results.

What Can You Do?

  1. Always follow the manufacturer's instructions for installing the infant seat in your vehicle and your infant within the seat. If your vehicle has UAS lower anchors, use them.
  2. Adjust the harnesses to fit your infant snugly each time you place him or her in the seat to ensure that he or she stays in the seat in a crash.

How research affects regulations

Laws require companies that make child seats to test them for safety when secured as normal. The seats must pass tests for sudden changes in speed of 48 km/h using a device called a sled. Only seats that meet the legal standards for safety can be sold in Canada.

Transport Canada research programs recently did over 50 in-vehicle crash tests. Since no law in the world requires these kinds of crash tests, why does Transport Canada do them? Because vehicle seats and child seat designs are always changing. We look for new areas of risk and decide if current regulations need to be updated to keep children safer than ever.

We use many vehicle models with different rear seat designs, different infant seats and different ways of installing them. Using child dummies that are about the size and weight of a 12-month old child, we crash the vehicles at different speeds and in different kinds of crashes. We also change the position of the vehicle's front seats and the size and weight of the adult dummies in the vehicle. We may even install child and infant seats next to each other, since this often happens in real life.

The types of crash tests we did

The infant seats were installed in the rear seats of vehicles undergoing four different kinds of frontal crash tests:

  1. frontal crash into a rigid barrier or wall at 40, 48 or 56 km/h

    Frontal crashes into a rigid barrier at speeds of 48 and 56 km/h are considered internationally to be severe crashes.
  2. frontal crash into an offset deformable barrier at 40 km/h
  3. frontal crash into the side of a stationary vehicle at 50 km/h
  4. frontal crash into the rear of a stationary vehicle at 50 km/h

Tougher standards for vehicles and child seats, as well as more police enforcement, have helped lower the number of children killed or hurt in vehicle crashes, even though more vehicles are on the road than ever before. Transport Canada will continue to work to ensure children are as safe as possible.

What we measured

Instruments on the test vehicles measured how quickly the vehicle changed speed during the crash. Instruments in the infant crash test dummy measured how quickly the dummy came to a stop and if the head or chest struck a hard surface during the crash. All results were studied using methods set by the Society of Automotive Engineers (SAE J-211).

The vehicles' rear doors were removed and replaced with beams so high-speed video cameras could get a complete view of the infant seat during the crash test.

How we installed infant seats

All but one seat were installed following the owners manuals' instructions. And while most tests used the seat and its base, three tests were carried out without the base.

The bases were attached either with the vehicle's lap and shoulder belt (often called a three-point belt) or with the lower anchors (also called the universal anchorage system or UAS). All seat bases were attached very tightly to the vehicle seat to reduce the amount of sideways movement. All vehicles were recent models and equipped with three-point belts in all seating positions. No in-vehicle tests used a two-point (lap only) seat belt found in some older vehicles.

In some tests, the driver and front passenger seats were placed as far to the front as possible, leaving a space between the front seat back and the infant seat, like this:

In other tests, the driver and front passenger seats were placed either in the middle or rearmost position, leaving little or no space between the infant seat and the front seat back, like this:

In one vehicle, we left the seat's handle in the carry position because there was not enough space for the handle to fit between the front and the rear seat. We know this is against the manufacturer's instructions, but we also know that some people may install their infant seats this way because they have no other way to install them.

How safe are the seats we tested?

Most seats did not separate from their base at 48 km/h. However, higher speed crashes could increase the chance of a seat separating from its base, as we learned during tests. Overall, the infant seats we tested provided excellent protection both in mild crash tests and in the most severe crash tests at 56 km/h.

During a frontal crash, the infant seats will rotate or swing towards the seat back. The video images may seem shocking at first, but this is in fact, the way they are designed to work. When the seat is correctly installed and the infant is securely restrained by the seat's harness, the infant remains in the seat shell and is very well protected from the violent forces of the crash. In severe crashes, the infant's face may come into contact with the top of the seat back they are facing or its headrest, but since this part of the seat is generally soft, the risk of injury is low.

Remember: Children are always safer in a seat that's properly installed and meant for a child of their age, size and weight.

Test results by method of attaching the infant seat

Since head contact can be difficult to measure simply from watching videos, instruments inside the crash test dummy head measure how quickly the infant's head changes speed and how hard it hits a surface (impact). What these tests show is that head impact greater than 80g, which could cause injury, was proven to occur in certain cases.

In the sample of 86 infant seat tests:

  • 57 infant seats attached to their base were secured with the vehicle lap and shoulder belt. In 10 cases, the infant dummy's head hit the front seat back with an impact of more than 80g.
  • 3 infant seats without the base attached were secured with the vehicle lap and shoulder belt. In one test, head impact was close to 80g, but did not go past it.
  • Only one of 22 infant seats secured with the lower anchors (UAS) resulted in head impact greater than 80g. This was due to contact with the handle and not because of the method of attachment.
  • One of the 4 convertible infant/child seats secured facing rearward with the lower anchors (UAS) resulted in the child dummy's head hitting the area between the head restraint and the shell of the seat.

We also did side-by-side tests to compare the lower anchors (UAS) and lap and shoulder belt attachment methods in the same vehicle. We attached two identical infant seats with bases behind the driver and the passenger seats of seven vehicles during a frontal wall crash at 48 km/h and in one vehicle during a wall crash at 56 km/h (the Eddie Bauer Surefit test). You can see the results of the eight paired tests in the chart below. It shows that:

  • In five tests, the head impact for infant dummies in the infant seat secured with the lap and shoulder belt was higher than those next to it in seats attached with the lower anchors (UAS).
  • In one test, the head impact for an infant dummy harnessed in a seat installed with the lap and shoulder belt was two times greater than the dummy in the identical seat next to it secured with the lower anchors.
  • In three tests (marked with *), head impact with the front seat back was seen on test videos.
  • In one test (marked 'SS') the Designer22 seat installed in the BMW X3 with the lap and shoulder belt separated from its base. Head impact occurred after separation. This prompted TC to issue a Consumer Information Notice.

Some seats separated from their base

In severe frontal crashes, three models of infant seats separated from their bases:

  1. Combi Center DX and ST

    The Combi seats separated from their bases in four of six vehicle crash tests. We used an acceleration sled to find out why the Combi Center and Safety 1st/Designer22 separated.

    The Combi Center DX and ST seats separated from their bases without any evidence of seat failure. In some crashes we discovered that there could be enough force to disconnect the seat from its base - as if you squeezed the release handle. The manufacturer issued a voluntary recall on February 28, 2008 to correct the problem.
  2. Safety 1st, Eddie Bauer, Cosco Designer 22

    Safety 1st, Eddie Bauer and Cosco Designer 22 infant seats separated from their bases in five of 22 crash tests. Further tests with an acceleration sled showed that the spring-loaded hooks used to secure the seat to a rod anchored in the base can swing open if the forces in the crash are high enough.

    During sled tests, we noted that while seats secured to the vehicle seat with the lap and shoulder belt could separate from the bases, seats that were secured with the UAS or only a lap belt stayed attached. The separation seemed to be caused by the shoulder belt lifting one side of the seat and creating a small gap between the hook and rod on one side. As a result, during a severe crash, the hook with the gap released and the seat moved enough to separate from both hooks.

    Transport Canada issued a Consumer Information Notice on August 5, 2009 recommending that consumers not use the base, but secure the Designer 22 seat directly to the vehicle using the three-point belt.
  3. Dorel Maxi-Cosi

    In one out of eight tests, the Dorel Maxi-Cosi infant seat separated from its base. The way this seat is secured to its base is different than the Combi Center and the Designer 22. We are doing more tests using the test sled to see if the seat separated because the seat was not installed correctly, or if there is another problem.

Crash severity is a factor

The severity of a crash depends on the speed at which the collision occurs and how the vehicle absorbs the energy of the crash. Remember, we used only some vehicle models that are common on Canadian roads. Vehicles that are able to slow down more gradually during a crash transfer less energy to a seat. This reduces the likelihood that a seat could separate from its base.

What else we learned

Some Designer 22 seats made from a different mould performed differently. Whether these seats were secured with the lap and shoulder belt, the lower anchors (UAS) or a lap only belt, in both vehicle and sled tests they developed cracks that extended from the top surface of the base, down the side and through the attachment point of the rod in the base.

What do our test results mean?

Most seats did not separate from their base at 48 km/h. Only two seats had a pattern of separating from their bases. However, higher speed crashes increased the chance of those seats separating from their bases. We bought all of the seats we tested from local stores. They all complied with Canadian regulations, but the required tests did not identify a risk of head impact or detect seat separation or base cracking. Why not? Child seats in Canada and the US are tested using what is called a bench seat (Figure 1) mounted on a sled that simulates the speeds and forces of a crash.


Figure 1: Bench seat used to conduct regulatory testing.

Differences between a bench seat used for compliance testing and in-vehicle seats include:

  • The seat cushion on the bench seat is softer than most seats found in vehicles today.
  • There is no seat in front of the test bench. That means, when a child seat is tested on the test bench, we can't observe head impact with a carry handle since there is no possible contact between the child seat and the back of the front seat. We can't detect head impact with a seat back, either.
  • The time it takes to drop from 48 km/h to 0 during a regulatory sled test is longer than the time it takes for many real motor vehicles to drop from 48 km/h to 0 during a rigid wall test. This means that the energy of the sled test crash for child seat regulations is less than the energy of many vehicle crashes into a rigid wall.
  • We tested infant seats installed with the lap and shoulder belt, which is sometimes the only option available. Why is this important? While regulations do not require this kind of testing, our research tests show that some seats installed with the lap and shoulder belt separate from their base.

Infant seat safety

Overall, the infant seats we tested provide excellent protection, even in severe crash tests. The tests show how violent a crash can be and how important it is to use an infant seat and to secure it tightly to the vehicle seat with the lower anchors (UAS), if you can. It is also very important to make sure that the seat harness fits snugly so your infant remains in the seat's protective shell.

A case for changing regulations?

Our tests show that infant seats perform differently in real vehicles than on a regulatory test sled. During a real crash, the infant seat will move towards the point of impact; if it is a frontal collision, that means the seat will move towards the front of the vehicle. As the infant seat moves forward it may strike the seat back of the driver and/or front passenger seat.

Existing regulations have really helped reduce the risk of death and injury for infants travelling in vehicles, but there is room to reduce that risk even more. Transport Canada takes the safety of children very seriously and we are always looking for ways to improve that safety by changing our regulations. Opportunities we identified for infant seats are:

  1. increasing the energy of the sled test used in the Canadian regulations;
  2. adding a requirement to test rear-facing infant seats installed with a lap and shoulder belt;
  3. using a standard test seat that is more like those found in the vehicles on the road today; and
  4. adding a physical barrier in front of the standard test seat that simulates the presence of a vehicle front seat.

Since the North American markets for children's restraint systems are so integrated, Canadian and US regulators are working together closely. Transport Canada is developing a proposal to update its regulation. Consultations with the U.S. regulators and North American children's restraint system manufacturers are ongoing, to help ensure that Canadians will have continued access to a broad selection of safe and affordable seat models. On August 15, 2009, Transport Canada published a Notice of Intent to amend its child seat regulations:

Transport Canada will continue to develop and issue notices to inform parents and caregivers of safety issues related to children's restraint systems as well as child road safety in general.