In crash testing it’s called time zero: the moment an accident begins. When we think about vehicle safety, we tend to think about what happens after time zero. Crumple zones engage.
Seat belts cinch tight. Airbags erupt. And after the violence ends, ideally the passenger cell remains intact, the humans inside unharmed. Those fractions of a second at the onset of an impact are crucial. But so are the ones that come before it. And the quest for safer cars runs in two directions – not just surviving a crash but trying to stop the clock before it ever gets to time zero.
For all the gains made in safety, getting into your car is probably the most dangerous thing you’ll do on any given day. On South African roads, about 14 000 people die annually in motor-vehicle accidents. But around the world car companies and governments are making advances in vehicle safety at an unprecedented pace.
In the past year I’ve visited the front lines of safety innovation, walking the floor of a crash-test facility in Ohio, talking to engineers across the globe, and trying out new electronic safety systems on the latest vehicles. From progress in active crash avoidance to huge improvements in materials, the achievements we’ve made in vehicle safety are staggering. We may never completely eliminate accidents or deaths, but we’re getting closer.
I. It starts with the steel
Considering all of the rapid developments we’ve seen with electronic safety systems in recent years, it’s perhaps counterintuitive that some of the biggest safety improvements in the past decade have come from good old-fashioned steel. “Over the past 10 to 15 years, steels have been getting stronger,” says Chuck Thomas, chief engineer at Honda R&D Americas, in Raymond, Ohio. “We probably had 500 megapascals of tensile strength in the early 2000s. Now hot-pressed or hot-stamped steel is around 1 500 megapascals.” At that strength you can hang around 100 000 kg on 3 cm-wide strip without tearing it in two. The high-strength steel is stamped hot and then quickly cooled, allowing for complex shapes and a wide variability in yield strength, which helps determine how a car deforms in an accident.
David Leone, executive chief engineer for Cadillac, says that the use of high-strength steel isn’t about turning passenger cars into invincible tanks but controlling crash energy and minimising weight. “Heavy does not mean safe,” Leone says. “Heavy means heavy. Go back to the ’50s and ’60s. The cars were heavy. They were stiff. But if you ran into the wall, you bounced off the wall and all the deceleration went through your body. Heavy and stiff is not where you want to be.”
These advances in steel – along with strategic use of other materials such as aluminium, magnesium and carbon fibre – allow engineers to design structures that can dissipate and redirect crash forces. For example, the new Cadillac CTS uses lightweight aluminium “crush cans” up front to soak up a lot of energy before an impact reaches the passengers. Even the CTS’s seat-belt spools unwind slightly during a crash to help minimise forces on your body.
The effective mix of stronger materials and crush zones is evident in a slow-motion video of the 2014 Acura MDX undergoing an offset-frontal crash test. As the car slams into the barrier at 64 km/h, the front end deforms alarmingly until the shock wave reaches the firewall, where it meets high-strength steel stamped at 870 degrees. Instead of continuing its collapse, the car pivots away from the barrier, absorbing the remaining energy. From the front door forward, the car is annihilated. From the door back, it’s completely intact.
II. We still need to smash stuff
“Don’t blink or you’ll miss it,” says a white-coated engineer as he prepares to fire the crash sled at Honda R&D Americas. The sled is fitted with a mockup of an Odyssey minivan dashboard, which is propelled backward by a 34 000-kW hydraulic piston. The dummy at the wheel will take a trip from zero to 56 km/h in 100 milliseconds – distance-wise, about 1,5 metres. The sled doesn’t actually crash into anything. The brute acceleration creates punishing g-forces that replicate those that occur in a front-end collision.
At the appointed moment the sled leaps backward and airbags deploy with a bang. The dummy tattoos the bag with paint dabbed on its head, the resultant smear telling a story of how a human might’ve fared in this hypothetical accident. This is one way Honda continues to improve its seat belts, airbags, seats, dash materials – items that don’t require the destruction of a whole car.
Of course, as with all other carmakers, Honda still runs full-scale tests in a cavernous clean room where actual cars are flung into various barriers under blinding lights, the impact recorded and dissected in super slo-mo. This is how carmakers test rollovers, side impacts, and small-overlap front-end collisions, which focus on the car’s right and left-front corners. The USA’s Insurance Institute for Highway Safety began running small-overlap tests in 2012 in response to statistics that roughly one in four front-end collisions involving serious or fatal injury fit this criteria – a car drifting into the oncoming lane catches the fender of an oncoming vehicle or goes off the road and clips a signpost.
These are particularly nasty impacts because they don’t engage the full crash structure of the front end. Instead, they tear through the vulnerable corners, sometimes forcing the front-left wheel into the driver’s footwell.
When the IIHS began evaluating small-overlap performance, the Volkswagen CC became the first car ever to have its driver-side door sheared completely off during a test. “If you don’t strike the columns on either side of the engine, that crash energy goes into the cabin,” Honda’s Thomas says. “We’ve done a lot of work to adapt to these kinds of crashes.”
In yet another room at Honda’s testing facility, mechanical arms launch plastic dummy heads into the interiors of two Acura RDXs. Every piece of the cabin is optimised to deform and cushion a blow, from the headliner to the plastic coat hangers above the windows. The R&D staff, who regularly witness the violence of car crashes, seem passively disdainful of people who don’t wear seat belts, the simplest and most effective way to prevent serious injuries in an accident. On the three test noggins are the names given to each dummy: Larry, Moe and Curly.
A full-size dummy named Polar II stands in for people during pedestrian-impact tests. He helped Honda develop a better design for its windscreen wipers. “If you look at your wipers, there’s probably a big bolt on the axis,” Thomas says. “Well, it turns out that’s a spot your head might hit and basically land on a big spike. The answer is a breakaway wiper system.” Which Hondas now use, thanks to Polar.
Built to crash
(see image and key above)
Carmakers use a variety of materials and steel strengths in a car’s frame to redistribute crash forces and protect passengers. In this 2015 Volvo XC90 there are five different grades of steel and lightweight aluminium. By using softer metals on the exterior parts and gradually using stronger steel through the crush zone and as part of the passenger cell, the violent energy from the impact can be controlled, keeping the humans inside safe. The XC90’s seven massive airbags will also help.
III. Vehicles are getting amazingly smart
Though crash performance is still critically important, much of today’s safety research concerns the relatively new field of active accident avoidance. “We know that collisions will still occur, so you have to work with the protection of the occupants,” Thomas Broberg, senior technical adviser for safety at the Volvo Cars Safety Centre, says. “That’s an evolution. The revolution, which has already started, is with collision avoidance – auto braking, steering, and autonomous driving.” Volvo was one of the first carmakers to market with autonomous emergency braking, which applies the brakes in certain situations if the driver does not. According to a study by the IIHS, Volvo XC60s equipped with the system were involved in 20 per cent fewer collisions than comparable SUVs without auto braking.
Self-steering cars are the next frontier. The past two years saw the introduction of self-steering by the Infiniti and Mercedes-Benz, each of which can make steering corrections at highway speeds to help maintain the car’s position in a lane. Currently these systems demand involvement from the driver; when the Mercedes-Benz determines that the driver hasn’t made a steering input in 16 seconds, the car shuts down its lane-keeping assistance. Lane keeping works to prevent inattentive drivers from drifting over the centreline or into the flank of an adjacent 18-wheeler, but that basic, limited functionality could soon be capable of taking over for long stretches of highway driving.
Vehicles that can brake on their own or steer themselves aren’t equipped with any one magic technology. Even small degrees of autonomy rely on a network of multiple sensors that are already used for the various active safety systems in a car, such as blind-spot or forward-collision alerts. Because each of these electronic systems relies on different technologies with different strengths, tying them together is the key to making a car semiautonomous and, maybe even one day, fully autonomous or driverless. “Cameras can distinguish shapes, but have difficulty judging distance and speed,” says Thomas. “Radar is good for distance and speed but not shapes. By using both together we get an accurate picture of the obstacles and what they are.”
Nissan claims it will offer an autonomous car by 2020. Audi’s piloted driving system, which can handle highway driving, is on track for release in three to five years. And in 2017, in Gothenburg, Sweden, 100 drivers will begin conducting their daily commutes in autonomous Volvos, part of a real-world research project that furthers Volvo’s stated goal of zero fatalities or major injuries in its cars by 2020. And, of course, Google continues to work on its driverless cars, which have covered so many kilometres without a major accident that states are scrambling to create laws to address this new category of vehicle.
There are plenty of reasons to be sceptical about how soon a production driverless car could hit public roads – legal and insurance issues being just two. However, automotive experts remain optimistic. “We’re going to look back in 15 or 20 years and say, that’s what used to be in cars? Remember when I had to steer on the freeways?” says John Capp, director of global vehicle safety for Cadillac. Capp led research and development on Cadillac’s upcoming Super Cruise system. Super Cruise takes over steering and pedal operations in certain highway conditions by using lane-keeping assistance paired with active cruise control, which together help a car maintain a set distance behind another vehicle without the driver having to apply the gas or brakes. These advanced systems blur the line between safety and luxury – when the car takes a share of stress away from the driver, safety moves to the foreground of your daily experience. “It’s not a lot of fun to try out your airbags, but using your active cruise control is,” Capp says.
IV. Our cars are learning to talk
As carmakers bring smarter cars to market, governments have been testing out technology that allows for vehicle-to-vehicle (V2V) communication. Greg Winfree is the assistant secretary for research and innovative technology with the US Department of Transportation, which announced plans in February to pursue getting the technology it’s been testing into production cars. “We’re working with car companies as we develop connected-vehicle technology,” Winfree says. “They recognise the safety potential, and they are as enthusiastic about it as we are.”
If cars can relay speed, braking, and position information to each other, then they’ll be able to register potentially hazardous situations almost instantaneously, warning the driver visually, audibly, or with the rumble of the seat or steering wheel. And if cars are equipped to brake and steer on their own, predictive accident avoidance becomes possible: your car could take action to avoid an unfolding situation that you can’t yet see.
“Because this is a connected system, all the vehicles must speak the same language,” Winfree says. This intercar chat will happen on the 5,9-GHz band of the radio spectrum, which the DOT and other international organisations have designated for transportation safety. This band allows communications between vehicles up to 10 times per second – a boon when cars on two-lane roads can approach each other at 90 km/h.
Eventually there is the potential for vehicles to talk to infrastructure such as stoplights. This could help traffic flow by co-ordinating the lights according to the situation. As for the obvious privacy question, fears that the government or nefarious individuals could use your car’s connectivity to track you, Winfree acknowledges that it’s a legitimate concern. To that end, these signals don’t transmit personal information, and the point of the programme isn’t to collect data. Besides, if someone wanted to data-mine you, there’d be easier ways to do it than by hacking a stoplight.
V. People are the problem
As our cars take on more driving tasks, there are suddenly entirely new engineering and design challenges for carmakers. “What is important as we go down this road, from the safety perspective, is how the car should interact with the driver,” Volvo’s Broberg says. “How do you let the car take over? How does the car tell you that you need to take over? We need to understand driver behaviour.”
The biggest hurdle is preventing drivers from over-relying on systems that aren’t intended to fully replace an alert human at the wheel. Subaru’s EyeSight auto-braking system will shut itself down after three consecutive near collisions, requiring the driver to restart the car before it resumes functioning. Audi’s piloted-driving hardware includes two cameras pointed back at drivers to determine if their eyes are closed for more than 10 seconds. Volvo is working on a similar concept that also scans the driver’s face with infrared lights to ascertain head position, making sure the driver isn’t nodding off.
Right now driver monitoring is used to determine the threshold of intervention for systems like lane keeping and auto braking. It is also used for basic yet helpful driver aids. In many Mercedes models a small indicator light in the shape of a coffee cup appears in the dash if the car determines the driver is getting tired. Eventually, with systems such as GM’s Super Cruise, this technology could tell the vehicle when to hand control back to the driver – or if a drowsy driver should even be allowed to activate the system in the first place. “It’s going to take a while before you can climb into the back seat and let the car take you to work,” Cadillac’s Capp says. “But we don’t have to wait for everything to be perfected before we take some steps in that direction.”
GM’s haptic seat, which vibrates to alert the driver of lane departure – and impending collisions – underscores the human-psychology role in safety. Studies revealed an unexpected benefit of the seat’s silent alert: drivers were more likely to use an electronic warning system if they could avoid the potential embarrassment of passengers knowing their mistakes. Sometimes pride goes before the fender bender.
As car technology continues to evolve, the only constant is the unpredictability of the people behind the wheel. Spotty seat-belt use, drinking and driving, texting – these are problems that the best engineers can’t vanquish. The good news is that behaviour and social mores can change just as fast as technology. Thirty years ago kids roamed free in the back of station wagons that lacked airbags, antilock brakes, or stability control. Now, not only are the cars themselves infinitely safer but an unbuckled child is an aberration rather than the norm. The world strives towards perfect machines and technologies that will minimise the dangers of a simple trip to work, school, or the grocery store, but the biggest variable in automotive safety is the same now as it was a hundred years ago. The final challenge, as always, is us.