Content Marketing Specialist, Novelda

Administrator's profile picture

Hot Car Deaths: Radar Technology Can Save Lives

On average, 37 children die in hot cars in the U.S. every year. That’s one every 9 days, more than 750 in 25 years1. And these shocking figures don’t mean that parents and childminders are being irresponsible - far from it.

Research shows that in 55% of cases of child heatstroke deaths in cars the responsible adult has unknowingly left their child in the car.2 Furthermore, in 28% of cases, a child has climbed into an unattended car on their own.

So in the vast majority of cases, these tragic deaths are not a result of any kind of deliberate negligence.

Anyone who’s been a parent knows what it’s like bringing up a young family. Sleep deprivation, the screaming, and the crying, balancing work with childcare. It’s not easy being a parent, it can be hectic, stressful and demanding, and human biology can sadly let us down.

‘How can you simply forget?!’ “How can you just leave a child in the car?!’ ‘How could you make a mistake like that?!’ The answer to such incredulous outbursts lies in the structure of our brains:

  • Information on what’s happening in our lives right now is stored in our hippocampus,
  • Information on future events, on changes to our daily routine are handled by our prefrontal cortex,
  • Information on habitual activities, those repeated routines we have reside in our basal ganglia.

And particularly when we’re very tired, the basal ganglia takes control and puts us in ‘auto-pilot’ mode. Our brain just does what it would normally be doing. Changes to our daily routine can be overlooked.

As a result, parents and child carers can simply become unaware that there’s a child in the back of a car. And it’s then that more than half of these tragic hot car deaths occur.

How hot does it need to get?

While the vast majority (87%) of children who die from heatstroke in cars are aged 3 or younger, children as old as 6 can also be victims. Temperatures in an unventilated car can sore and children’s bodies aren’t equipped to cope.

At body temperatures above 40ºC (104ºF) human organs are at risk, and since childrens’ bodies can overheat three to five time faster than adults’, they can be in danger very quickly.

Bear in mind that two-thirds of the increase in temperature inside a car can happen in under 20 minutes and that cracking a car window will not help either slow the rise in temperature or decrease the top temperature, and the peril a child is in will be clear. As a result of these factors, children have died from heatstroke in cars where the outside temperature has been as low as 15.5ºC (60ºF). So no, this life-threatening problem is not restricted to the U.S., where average temperatures are higher say than here in Europe. Consider for a moment a much more temperate North European climate like that in the United Kingdom for example. The mean summer temperature in the country was 14.9ºC (59ºF) in 2016 with a peak temperature of 21ºC (70ºF).

And yes, while the reported cases of children dying from heatstroke in cars are certainly less in Europe than it is in the U.S., there have been fatalities nevertheless, most notably in Belgium and France.

The interior temperature of a stationery, unventilated car can rise 10ºC in just 10 minutes, wherever it is in the world. It’s an international danger.

How progress hasn’t helped

There’s also a tragic irony that our valid concerns for the health and safety of our children in cars have led to advances that have unwittingly not helped the situation.

The car industry’s rightful introduction of airbags to protect front seat passengers in the event of a road traffic accident meant that we moved children’s car seats from the front of the car to the back – out of direct line of sight.

And research figures 3 show that during the 1990s child deaths due to airbags diminished from a peak of more than 60 per year to zero, while the number of deaths due to heat stroke gradually rose to reach the average annual figure of 37.

What’s more, the car industry also rightly concluded that rear-facing childrens’ car seats were much better at protecting our children in the event of a car crash. In most cases what this means is that our children (particularly babies) aren’t even visible from the front of the car.

So ironically, as a result of good progress in health and safety measures in cars, we risk being even less aware of our children in the back of a car.

Changing human behavior

In the absence of ready technical solutions to the problem, campaign groups have advocated a host of good measures to help us change our behavior and ensure we remember that a child is in the back seat.

These reminders include:

  • Putting something you really depend on in front of the same seat as your child, such as a mobile phone, a handbag or a wallet.
  • Keep a big soft toy in the children’s car seat and move it to the front seat when you put your child in the car.
  • Place the children’s car seat behind the front passenger seat where it’s more visible to the driver.

And there are a lot more suggestions than these. The challenge though is that we’re just not infallible, we’re human and when you’re a tired, stressed, frazzled parent mistakes can still happen.

Finding a technical solution

So how difficult can it really be though to find a technical solution to the problem? After all, our cars already remind us to put a seatbelt on, tell us that a car door is open or our lights are on, informs us when we’re low on petrol. So why can’t it remind us that we have a child in the back?

The automotive industry has begun to react, with a system for example that reminds drivers on arrival at a destination that they need to check the back seat. It remembers that a back door was used on a journey start and alerts the driver with a suitable dashboard message and tone.

And what about incorporating a back seat weight-triggered sensor? Just take the same sensor technology that tells a driver to disable the airbag when a small child is sitting in the front seat and apply it to the back seat. Or integrate a suitably sited proximity sensor instead perhaps.

Or maybe build a system around the seat belt buckle sensors. After all, they’re already in place and tell us faithfully when they aren’t being used.

Whatever technical solutions are developed in the future, above all else they need to be wholly accurate and reliable at all times. Whether a child is aged 1 or 6, whether the child is asleep or awake, in a children’s car seat or not, it needs to reliably cover all eventualities.

Remember that in almost 30% of cases of car heatstroke deaths, children have entered the car themselves and got locked in. They might be in the front seat, not the back, and they might not be sat in a seat at all.

At Novelda, we believe we now have the technology to enable ultra-reliable solutions to be developed and added to our cars to help protect against child heatstroke deaths in cars.

Why radar can work

The challenge in detecting the very smallest human movements in an enclosed space has been that traditional non-contact occupancy sensor technology just doesn’t have the necessary capability.

Passive infrared (PIR) sensors will detect what we call ‘major motion,’ like a person walking. Ultrasonic sensors are more sensitive, and will also detect what is called ‘minor motion’, a hand movement for example.

However, when we’re talking about a little baby fast asleep, almost motionless in a rear-facing car seat in the back of a car, a sensor needs to be able to detect what Novelda is calling ‘finer motion’.

A child breathing exhibits movement, and occupancy sensor technology needs to be capable of detecting this if an infant presence sensor for cars is to operate reliably and effectively.

What about Doppler radar?

Although not as widespread in use, Doppler radar-based occupancy sensors address many of the limitations of both PIR and ultrasonic non-contact occupancy sensors. Indeed the base technology has been used to monitor finer physiological movements for some 50 years.

The problem has been, however, that microwave solutions using the Doppler Effect have been ‘continuous wave’ and narrowband in nature, meaning they transmit a known frequency and receive the bounce backs from reflecting objects all of the time. This again needs a significant amount of power, and mostly analog designs have been rather costly to deploy.

What’s more, while radar signals will certainly pass through clothing materials and are sensitive enough to detect repetitive cardiopulmonary movements, they only provide phase information and just can’t measure absolute distance or position in a car.

As a result, even continuous wave radar-based occupancy sensors can’t distinguish between simultaneous finer human movements and deliver the absolute reliability that an infant presence sensor demands.

A solution using UWB pulse radar?

In arriving at an occupancy sensor capable of reliably detecting major, minor and finer human motion, Novelda focused its attention on a radar-based solution, and set itself the target of resolving the known distance measurement, power and BOM cost deficiencies of existing technologies.

Instead of continuous wave narrowband operation, ultra-wideband (UWB) pulse radar techniques for emitting and sampling signal pulses were proven to achieve highly accurate absolute distance measurements, determined by the time differences between transmitted and received pulses.

And by taking what is essentially a spread spectrum approach and using digital signal processing to recover the return signal, UWB radar can operate at much lower power levels than conventional radar.

This then overcomes consumer concern of not wanting a high-power radar sitting within touching distance of a child. To give it some context, our technique allows radar operation at power levels less than 1/1000th that of a regular Bluetooth headset.

Radar occupancy sensing ‘out-of-the-box’

Launched in 2017, Novelda’s X4M300 occupancy sensor, illustrated in Figure 1, integrates all of the components (and more) required by an infant presence monitor.

X4M300 Presence Sensor
Figure 1. Novelda’s UWB impulse radar presence sensor.

It comprises the XeThru X4 radar SoC, PCB antennas, a microcontroller, glue logic, PSU, USB interface, GPIO and LED status indicators. Its BOM cost is low, its physical size is small, its operating temperature seriously industrial. It’s ready to go.

Able to deliver sub-mm sensing accuracy at distances up to 10m means the module provides the reliable major, minor and finer motion detection capability that car equipment manufacturers need from a car occupancy sensor:

  • Extreme sensitivity – detects a child breathing
  • Sees through – clothes, blankets, seats
  • Measures range – distance to the child
  • Long life – no recalibration needed, no dust or dirt sensitive lenses
  • Hidden – inside a door, seat or dashboard
  • Non-intrusive – it isn’t a camera
  • Safe – extremely low emission levels

What’s more, the X4M300 occupancy sensor module single-handedly ‘debunks’ radar. It takes care of the radar, the antenna, the PCB and signal processing algorithm design. There is a complete hardware and software development kit too (the X4M03). The hard work has been done.

So radar is no longer the preserve of complex and costly, high-end markets, it's ready for 100% reliable infant presence detection in cars. Whether our children’s movements are major, minor or finer, they’ll be detected and the driver can be automatically reminded: ‘Don’t forget, you have someone very precious in the back seat of your car.’

Read more about the X4M300 presence sensor

To find out more about hot car deaths, I would strongly recommend that you visit, a charitable organization in the U.S. that’s committed to keeping children safe in and around cars.

(2) National Highway Traffic Safety Administration (NHTSA)
(3) National Center for Statistics and Analysis (NCSA), NHTSA