There’s always a small lag between when a driver turns a steering wheel and when the car actually responds. This lag causes the person at the helm to make numerous tiny corrections—corrections that Mazda wants to reduce with G-Vectoring Control, a new system that uses clever software to shift weight onto the front tires during turn-in. It works, but it’s subtle. Maybe too subtle.
Last week when I first wrote about Mazda’s G-Vectoring Control, I said the system “hints at a greater focus on comfort,” but after actually driving vehicles equipped with GVC, I’ve realized that there’s much more to Mazda’s software trickery than creating a more cushy ride quality.
[Full disclosure: Mazda needed me to try this new technology so badly they flew me out to San Francisco, drove me to Monterey, and handed me the key to a hotel room. The following morning, they drove me to Laguna Seca and trusted me with keys to their brand new automobiles. I behaved.]
In Theory
Mazda’s goal when developing G-Vectoring Control was to give its cars a behavior that minimizes jerk, the technical term for a change in acceleration over time—like that kick in the ass you get when your turbo kicks in. The company says the human body naturally minimizes jerk when it moves, so creating a car with similar behavior should help the car feel like more of “an extension of your own body,” Mazda says.
G-Vectoring Control reduces lateral jerk in corners by minimizing the transient response time between steering input and lateral acceleration. In other words, the time it takes the car to do what you want it to is reduced.
This more linear steering response means the driver doesn’t have to steer too far to correct for the initial lag, only to then have to correct again once the car “hooks up,” sending the car into sinusoidal, jerky motion.
This makes the steering feel more direct, and improves the feedback loop between the car and the driver, reducing swaying that is so prevalent in turns, particularly on rough or wet roads. In theory, it could make a car safer under these conditions. But only a tiny bit.
To reduce this delay between steering input and vehicle response, Mazda worked with Hitachi to create a system that automatically shifts weight to the front tires during turn-in to load up the sidewalls and suspension (this is one of the benefits to “left foot braking,” which is done often in racing).
But GVC doesn’t use brakes, it actually reduces vehicle acceleration and loads up the front tires by pulling spark timing. Mazda’s system rotates the vehicle about its center of mass and exerts weight on the front tires by actually reducing engine torque—a very unique approach that Mazda says is an industry first.
Mazda describes the method thusly:
GVC maximizes tire performance by focusing on the vertical load on the tires. The moment the driver starts to turn the steering wheel, GVC controls engine drive torque to generate a deceleration G-force, thereby shifting load to the front wheels. This increases front-wheel tire grip, enhancing the vehicle’s turn-in responsiveness.
How much torque is reduced depends on the rate of change of steering angle. If you crank hard on that wheel, for example, the system pulls more torque to load the wheels up faster. If you stop turning and keep your wheel at a steady position, engine torque returns to its initial value, shifting weight rearward for better stability.
The system is triggered only when a driver’s foot is on the throttle and the steering wheel is actively being turned. It does not work when the driver is on the brakes or when the steering wheel is not moving.
Ultimately, Mazda says, by reducing torque and shifting weight to the front wheels during turn-in, the system allows the car to move “more precisely as the driver intends, reducing the need for steering corrections, many of which are performed unconsciously.”
In Practice
Mazda began its presentation by admitting that the benefit of the system is “subtle but important.” And boy are they right on the “subtle” part, as I could hardly tell a difference between GVC on and GVC off. That’s no surprise, though, as within 50 milliseconds of steering input, GVC is only pulling a tiny bit of torque (in the single digit lb-ft), and causing a change in longitudinal acceleration of only 0.01 to 0.05G.
But that’s apparently enough to shift up to 10 pounds of weight to the front tires, which Mazda says creates a small benefit in steering response, which then creates a large reduction in drivers’ steering corrections, resulting in a smoother cornering experience .
Mazda had me drive a Mazda 6 sedan with a giant red button on the infotainment screen that turned G-Vectoring Control on and off. I did lots of different kinds of driving, including a moose test, a wet handling course, and an excruciatingly-boring 20 mph lap around Mazda Raceway Laguna Seca.
During most of these, I couldn’t really tell much of a difference with the system on or off. One section of that 20 mph lap around Mazda Raceway, though, required me to squeeze the sedan between some orange cones and a barrier, and that’s where I think I noticed a slight change.
With the system on, I easily made it through the narrow opening without breaking a sweat. With GVC off, I found myself tip-toeing through, and then absolutely annihilating a cone (luckily, not the wall.)
Did I eat the cone with the system off because the initial steering response time was too long, causing me to steer too far, only then to over-correct and eat a cone? Was GVC actually working? Apparently so, based on the data Mazda showed me. My steering angle amplitudes were much higher with the system off, indicating that I was making much bigger corrections.
That poor cone.
So it seems like the system works, even if I could rarely tell the difference. Other more “perceptive” journalists did say they felt a distinct difference, particularly on the wet handling course. And occupants also said that, with GVC on, the reduction in steering corrections made for a significantly more relaxed ride with less swaying side-to-side.
Honestly, 99.9 percent of customers aren’t even going to notice the difference GVC makes on steering response or ride comfort. But considering it’s a purely software-based system that adds no additional weight, and the fact that Mazda’s data shows a reduction in steering corrections, this seems like a win to me.
G-Vectoring Control will debut in the 2017 Mazda6 as standard equipment, and will make its way into the rest of the Mazda lineup sometime down the road.
It would be interesting to see how much benefit such a system would have on a big, sloppy SUV as well, like a soft-sprung Jeep Wrangler. That thing could use as much help in the “steering directness” department as it can get.